Ipc J-Std-030a-2014

I PC J -STD-030A
Febru ary 2 01 4
Su persed es J -STD-030
JOINT
Septem ber 2 005
INDUSTRY
STANDARD
Selection and
Application
of Board Level
Underfill Materials
Copyright IPC-Association Connecting Electronics Industries

The Principles of In May 1 995 the IPC’s Technical Activities Executive Committee (TAEC) adopted Principles of
Standardization S tandardization as a guiding principle of IPC’s standardization efforts.
Standards Should: Standards Should Not:

• S how relationship to Design for Manufacturability • Inhibit innovation
(DFM) and Design for the Environment (DFE) • Increase time-to-market
• Minimize time to market • Keep people out
• Contain simple (simplified) language • Increase cycle time
• Just include spec information • Tell you how to make something
• Focus on end product performance • Contain anything that cannot
• Include a feedback system on use and be defended with data
problems for future improvement
Notice IPC S tandards and Publications are designed to serve the public interest through eliminating mis-
understandings between manufacturers and purchasers, facilitating interchangeability and improve-
ment of products, and assisting the purchaser in selecting and obtaining with minimum delay the
proper product for his particular need. Existence of such S tandards and Publications shall not in
any respect preclude any member or nonmember of IPC from manufacturing or selling products
not conforming to such S tandards and Publication, nor shall the existence of such S tandards and
Publications preclude their voluntary use by those other than IPC members, whether the standard
is to be used either domestically or internationally.
Recommended S tandards and Publications are adopted by IPC without regard to whether their adop-
tion may involve patents on articles, materials, or processes. B y such action, IPC does not assume
any liability to any patent owner, nor do they assume any obligation whatever to parties adopting
the Recommended S tandard or Publication. Users are also wholly responsible for protecting them-
selves against all claims of liabilities for patent infringement.
IPC Position It is the position of IPC’s Technical Activities Executive Committee that the use and implementation
Statement on of IPC publications is voluntary and is part of a relationship entered into by customer and supplier.
Specification When an IPC publication is updated and a new revision is published, it is the opinion of the TAEC
Revision Change that the use of the new revision as part of an existing relationship is not automatic unless required
by the contract. The TAEC recommends the use of the latest revision. Adopted October 6, 1 998
Why is there Your purchase of this document contributes to the ongoing development of new and updated industry
a charge for standards and publications. S tandards allow manufacturers, customers, and suppliers to understand
this document? one another better. S tandards allow manufacturers greater efficiencies when they can set up their
processes to meet industry standards, allowing them to offer their customers lower costs.
IPC spends hundreds of thousands of dollars annually to support IPC’s volunteers in the standards
and publications development process. There are many rounds of drafts sent out for review and
the committees spend hundreds of hours in review and development. IPC’s staff attends and par-
ticipates in committee activities, typesets and circulates document drafts, and follows all necessary
procedures to qualify for ANS I approval.
IPC’s membership dues have been kept low to allow as many companies as possible to participate.
Therefore, the standards and publications revenue is necessary to complement dues revenue. The
price schedule offers a 50% discount to IPC members. If your company buys IPC standards and
publications, why not take advantage of this and the many other benefits of IPC membership as
well? For more information on membership in IPC, please visit www. ipc. org or call 847/597-2872.
Thank you for your continued support.
©Copyright 201 4. IPC, Bannockburn, Illinois, USA. All rights reserved under both international and Pan-American copyright conventions. Any
copying, scanning or other reproduction of these materials without the prior written consent of the copyright holder is strictly prohibited and
constitutes infringement under the Copyright Law of the United States.
I PC J -STD-030A
Selection and Application

of Board Level Underfill
Materials
Developed by the Underfill Materials Design, Selection and Process Task

Group (5-24f) of the Assembly and Joining Committee (5-20) of IPC
Supersedes: Users of this publication are encouraged to participate in the

J-STD-030 - September 2005 development of future revisions.
Contact:
I PC
3 000 Lakeside Drive, S uite 3 09S
B annockburn, IL 6001 5-1 249
Phone ( 847) 61 5-71 00
Fax (847) 61 5-71 05

This Page Intentionally Left Blank

February 201 4 IPC J-STD-030A
Acknowledgment
Any document involving a complex technology draws material from a vast number of sources. While the principal members
of the Underfill Materials Design, S election, and Process Task Group (5-24f) of the Assembly and Joining Committee (5-20)
are shown below, it is not possible to include all of those who assisted in the evolution of this standard. To each of them,
the members of the IPC extend their gratitude.
Assembly and Joining Underfill Materials Design, Technical Liaison of the

Committee Selection and Process IPC Board of Directors
Task Group
Chair Co-Chairs
Leo P. Lambert B ill R. Vuono B ob Neves
EPTAC Corporation TriQuint S emiconductor Microtek Laboratories
Vice Chair B rian J. Toleno

Renee J. Michalkiewicz Henkel Corporation
Trace Laboratories - B altimore
Vice Chair
Fonda B . Wu
Raytheon Company
Underfill Materials Design, Selection and Process Task Group

Joel Peiffer, 3 M Company Jagadeesh Radhakrishnan, Intel Nate Grinvalds, Rockwell Collins
Constantino Gonzalez, ACME Corporation David Hillman, Rockwell Collins

Training & Consulting Venmathy Raj arathinam, Intel E. Hofer, Rockwell Collins
Karl S eelig, AIM, Inc. Corporation
Douglas Pauls, Rockwell Collins
Randy Temple, Alltemated, Inc. Adrian Hawkins, Kester
Debie Vorwald, Rockwell Collins
Tek S ing Ong, Alpha David Lober, Kyzen Corporation
Greg Hurst, RS I, Inc.
Karen Tellefsen, Alpha Dock B rown, Medtronic
Fred Rosen, S andia National Labs
Agnieszka Ozarowski, B AE S ystems B rian S chmaltz, Namics
Albuquerque
B everley Christian, B lackB erry Technologies, Inc.
Yanmei Li, S chlumberger Well
Manny S anchez, B oeing Company Dan Ashley, Nordson AS YMTEK
S ervices
Michael S olis, B oeing Company Vicki Worth, Nordson AS YMTEK
Kelly S chriver, S chriver Consultants
S tan Rak, Continental Automotive Wennei Chen, Northrop Grumman
S arah Zarrin, S eagate Technology
GmbH Aerospace S ystems
Vern S olberg, S olberg Technical
Laura Cohen, Continental Automotive Mahendra Gandhi, Northrop
Consulting
S ystems Grumman Aerospace S ystems
S teven Martell, S onoscan Inc.
Mark Fulcher, Continental Gustavo Arredondo, Para Tech
David Hwang, S pace Exploration
Automotive S ystems Coating Inc.
Technologies
Virgil S hort, Eastman Kodak Mumtaz B ora, Peregrine
Michael Firmstone, Thermoset - Lord
Company S emiconductor
Chemical Products
Julie Filips, Elbit S ystems of America Richard Kraszewski, Plexus Corp.
Debora Obitz, Trace Laboratories -
Ken Gilleo, ET-Trends LLC Kenneth Manning, Raytheon
B altimore
Paco S olis, Foresite, Inc. Company
B ill Vuono, TriQuint S emiconductor
Hector Pallavicini, Harris Corporation David Nelson, Raytheon Company
Neil Murray, TRW/Automotive
George Carson, Henkel Corporation Jeff S eekatz, Raytheon Company
Electronics Group
B rian Toleno, Henkel Corporation Jeff S hubrooks, Raytheon Company
Tzyy Jang Tseng, Unimicron
Raiyomand Aspandiar, Intel Royce Taylor, Raytheon Company
Technology Corp.
Corporation Fonda Wu, Raytheon Company
Joe Fj elstad, Verdant Electronics
Kayleen Helms, Intel Corporation Martin S cionti, Raytheon Missile
Jim Watson, ZiiLabs
Ife Hsu, Intel Corporation S ystems
Karl Loh, Zymet Incorporated
James Maguire, Intel Corporation Caroline Ehlinger, Rockwell Collins
iii
IPC J-STD-030A February 201 4
iv
Table of Contents
1 SCOPE ...................................................................... 1 7. 2 S torage Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0
1 .1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7. 3 Preconditioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1
1 .2 Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7. 4 Pot Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1
1 .3 Definition of Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7. 4. 1 Viscosity Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1
7. 4. 2 Flow Rate Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1

2 APPLI CAB LE DO CU M E N TS .................................... 2
7. 4. 3 S ettling Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1
2. 1 IPC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. 2 American S ociety for Testing and Materials 8 APPLI CATI ON PR OCESS ...................................... 11
(AS TM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
8. 1 Pre-Application B oard Preparation . . . . . . . . . . . . . . . . . . 1 2
2. 3 Telcordia Technologies, Inc. .............................. 2
8. 2 Application of Capillary Flow Underfill . . . . . . . . . 1 2
3 TE RM S AN D DE FI N I TI O N S ...................................... 2 8. 2. 1 Dispensing Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3
8. 2. 1 . 1 Dispensing Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3
4 B ACKG RO U N D ......................................................... 3
8. 2. 1 . 2 Process Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 4
4. 1 Why Is Underfill Needed? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
8. 2. 2 Application Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 4
4. 2 Types of Underfill . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
8. 2. 2. 1 Air Entrapment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 4
4. 2. 1 Capillary Underfill . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
8. 2. 2. 2 Gravitational Phase S eparation . . . . . . . . . . . . . . . . . . . . . . . . 1 4
4. 2. 2 Fluxing (No-Flow) Encapsulant . . . . . . . . . . . . . . . . . . . . . . . . . 4
8. 2. 2. 3 Dynamic Phase S eparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5
4. 2. 3 Removable/Reworkable Underfill . . . . . . . . . . . . . . . . . . . . . . 5
8. 2. 2. 4 Filtering Phase S eparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5
4. 2. 4 Cornerbond Adhesive (Dispensed or
Placed Prior to Reflow Process) . . . . . . . . . . . . . . . . . . . . . . . . 5 8. 3 Application of No-Flow/Fluxing Underfill . . . . . . 1 5
4. 2. 5 Corner Glue/Corner Tack Underfills 8. 3 . 1 Dispensing Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5
(Applied After S MT Reflow) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 8. 3 . 2 Dispense Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5
8. 3 . 3 Package Placement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5
5 M E CH AN I CAL CO N SI DE RATI O N S ......................... 6
5. 1 Footprint Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 8. 3 . 4 Application Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 6
5. 2 Gap S ize . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 8. 4 Flow Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 6
5. 3 Pad Redistribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 8. 4. 1 Dispense Flow Rate Measurement . . . . . . . . . . . . . . . . . . 1 6
8. 4. 2 Underfill Flow Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 6

6 U N CU R E D U N DE RFI LL CH AR ACTE R I STI CS ........ 7
8. 4. 3 Flow Out and B leed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 6
6. 1 Filler Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
8. 5 S pread/S lump . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7
6. 1 . 1 Filler S ize . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
8. 6 Evaluation Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7
6. 1 . 2 Filler Material Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
8. 6. 1 Acoustic Micro-Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7
6. 1 . 3 Filler Content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
8. 6. 2 Destructive Tests for Voiding . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7
6. 1 . 4 Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
8. 6. 3 Assembly to Glass for Flow Visualization . . . . . . 1 8
6. 2 Prepolymer Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
8. 7 Pot Life (In Dispenser) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8
6. 2. 1 Viscosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
6. 2. 2 Gel Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 9 CU R E PRO CE SS .................................................... 1 8
6. 3 Material Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 9. 1 Applied Life (After Dispensing) . . . . . . . . . . . . . . . . . . . . . . 1 8
6. 3 . 1 Flux Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 9. 2 Cure Process for Capillary Flow Underfill . . . . . . 1 9
6. 3 . 2 B oard S urface Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 9. 2. 1 Process Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 9
6. 3 . 3 Component S urface Compatibility . . . . . . . . . . . . . . . . . . . . . 9 9. 2. 2 Cure S chedule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 9
6. 4 Alpha Particle Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 9. 2. 3 Heating Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 9
9. 2. 4 Temperature S ensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 9
7 M ATE RI ALS PACKAG I N G , H AN DLI N G , AN D
9. 3 Cure Process for No-Flow Underfill . . . . . . . . . . . . . . . . 1 9
STOR AG E ............................................................... 1 0
7. 1 Packaging and Containers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0 9. 4 Void Formation/Outgassing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
7. 1 . 1 Voids/B ubbles in Packed Material . . . . . . . . . . . . . . . . . . . 1 0 9. 5 Cure Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
v
1 0 CU RE D U N DER FI LL CH ARACTER I STI CS . . . . . . . . . 20 1 3.6 General Overview of Voiding Conditions
1 0. 1 Appearance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 and Reliability Impact/Factors . . . . . . . . . . . . . . . . . . . . . . . . . 29
1 0. 1 . 1 Fillet Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 1 3.7 Workmanship Guidelines for Corner or

Edge B ond Underfills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
1 0. 1 . 2 Color (Dye/Pigment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1 0. 2 Adhesion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 1 4 TRO U B LESH OOTI N G .......................................... 30
1 0. 3 S hrinkage and Induced S tress . . . . . . . . . . . . . . . . . . . . . . . . . . 21 1 4. 1 Inadequate Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0
1 0. 4 Young’s Modulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 1 4. 1 . 1 Viscosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0
1 0. 5 Coefficient of Thermal Expansion (CTE) . . . . . . . . 21 1 4. 1 . 2 Wetting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0
1 0. 6 Glass Transition Temperature (Tg ) . . . . . . . . . . . . . . . . . . . 21 1 4. 1 . 3 Mechanical B lockage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0
1 0. 7 Chemical S tability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 1 4. 2 Phase S eparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0
1 0. 7. 1 Determining Resistance to S olvents . . . . . . . . . . . . . . . . 21 1 4. 3 Voids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0
1 0. 8 Moisture Absorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 1 4. 3 . 1 Voids B efore Cure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0
1 0. 9 Hydrolytic S tability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 1 4. 3 . 2 Voids After Cure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0
1 0. 1 0 Non-Nutrient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 1 4. 4 Inadequate Cure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0
1 0. 1 1 S urface Insulation Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . 22 1 4. 5 Poor Adhesion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1
1 0. 1 2 Electrochemical Migration Resistance . . . . . . . . . . . . . 23 1 4. 6 Thermal Cycle Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1
1 0. 1 3 Volume Resistivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Fi g u res
1 0. 1 4 Permittivity (Dielectric Constant) . . . . . . . . . . . . . . . . . . . . 23

Figure 1 -1 Comparison of Various Sized Array
11 WOR KM AN SH I P . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Packages ........................................................... 1
11 .1 S ubstrate Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Figure 4-1 Both the Flip-Chip and CSP Underfill in
a Flip-Chip CSP Soldered to a PCA ................. 3
11 .2 Cleaning B efore Underfill . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Figure 4-2 Different Material Coverage Types for
11 .3 Cleaning After Cure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Top and X-Sec View (Underfill, Corner
Glue, Epoxy Flux) .............................................. 3
1 2 PRO CE SS RE LI AB I LI TY ASSE SSM E N T . . . . . . . . . . . . 24
Figure 4-3 Example of Epoxy Chemical Reaction .............. 4
1 2. 1 Ionic Content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Figure 4-4 Image of Needle Dispensing of an
1 2. 2 Chemical Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Underfill (Bottom Side View) ............................. 4
1 2. 3 Mechanical Integrity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Figure 4-5 Fluxing Underfill Process .................................. 5
1 2. 5 Post S oldering Processes (Capillary
Figure 4-6 Reflow Cured Corner Bond Process ................. 6
Underfill) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Figure 4-7 Thermally and UV Cured Corner Tacks ............ 6
1 2. 6 Temperature Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Figure 6-1 SEM Image Showing Poor Underfill
Adhesion to a Bump Due to Flux
1 3 O TH E R CO N SI DE RATI O N S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Residue ............................................................. 9
1 3.1 Reworkability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Figure 8-1 Fluxing Underfill ............................................... 1 2
1 3.1 .1 Rework of Adj acent Components . . . . . . . . . . . . . . . . . . . . 26
Figure 8-2 Underfill Dispensing ........................................ 1 3
1 3.2 Determination of Cure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Figure 8-3 Examples of Dispensing Patterns
(The Fillet is Shown in a Different
1 3.3 Thermal Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Color for Clarity Only) ...................................... 1 3
1 3.4 Workmanship S tandards for B GA and Figure 8-4 Image Showing Both Needle Dispensing
CS P B oard Level Underfills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 and Jetting for Underfilling at the Chip
Level ................................................................ 1 4
1 3 . 4. 1 Visual Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Figure 8-5 Air Entrapment (Underfill was Dispensed
1 3 . 4. 2 Underfill Voiding Under Devices . . . . . . . . . . . . . . . . . . . . . 28 on the Left Edge of the Package) ................... 1 4
1 3.5 Destructive Inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Figure 8-6 Gravitational Phase Separation ...................... 1 4
vi
Figure 8-7 Dynamic Phase Separation ............................. 15

Figure 8-8 Filtering Phase Separation .............................. 15
Figure 8-9 Underfill Bleed ................................................. 17
Figure 9-1 Filler Settling within Underfill ........................... 18
Figure 1 0-1 Underfill Adhesion Test .................................... 21
Figure 1 2-1 Material Property and Thermal Cycling
Response ........................................................ 25
Figure 1 3-1 Acceptable – Class 1 ,2,3 ................................ 27
Figure 1 3-2 Void Comparison ............................................. 28
Figure 1 3-3 Target – Flat Cross Section that Shows
No Voids in the Underfill Layer ....................... 29
Figure 1 3-4 Image of a Component with Edgebond
Underfill ........................................................... 29
Figure 1 3-5 Fillet Width ....................................................... 29
Figure 1 4-1 Filtering Phase Separation .............................. 30
Tables
Table 4-1 Guidelines for Evaluation .................................... 4

Table 8-1 Test Methods ..................................................... 1 8
Table 1 0-1 Conditions for Chemical Resistance Testing .... 22
vii
viii
Selection and Application of

Board Level Underfill Materials
1 SCOPE
This document provides users of underfill material with guidance in selecting and evaluating underfill material for assem-
bly solder j oints second level interconnects. Underfill material is used to increase reliability of electronic devices by two
methods: alleviate coefficient of thermal expansion (CTE) mismatch (between the electronic package and the assembly sub-
strate) and/or increase mechanical strength. Underfill materials are also used for environmental protection, mechanical shock
or vibration, and anti-tampering uses. Materials used in underfill applications should not adversely affect device reliability
nor degrade electrical performance (e. g. , ionic impurities). When correctly selected and applied, underfill material should
increase the life of the assembly solder j oints.
Types of underfill materials currently available in the market include:
• Capillary Flow Underfill
– Primary UFs (Package level not within the scope of this document)
– S econdary (B oard level)
• No-Flow/Fluxing Underfill
– Thermal Compression B onding (TCB ) Epoxies (not within the scope of this document)
– Non-Conductive Paste (NCP)
– Non-Conductive Film (NCF)
• Removable/Re-Workable Underfill
• Corner B onding/Glue B onding
• Molded Underfill (not within scope of document)
• Wafer Applied Underfill (not within scope of document)
• Vacuum Underfill (not within scope of document)
1 .1 Introduction This document covers polymer based underfill materials intended for use in printed circuit assemblies
(PCA). (S ee Figure 1 -1 . )
1 .2 Purpose The purpose of this document is to help in

identifying underfill materials whose properties are compat-
G
ible with c omp onent as s emb ly j oints to reduce thermo
mechanical stress so that performance of the assembly is
enhanced. The additional role of underfill is protecting the A
device from environmental factors and increasing strength. D
Evaluation methods are provided in the document that are
H
B
intended to be used for assessing underfill material perfor-
mance in specific applications as well as troubleshooting fail-
ures. This document represents the compiled knowledge and
E
experience of the IPC Underfill Adhesives for Flip Chip
Applications Task Group.
C F
IPC-030a-1 -1
1 .3 Definition of Requirements The word ‘‘ shall ’’ is used
in the text of this document wherever there is a requirement Figure 1 -1 Comparison of Various Sized Array Packages
for materials, preparation, process control, or acceptance of a A. Ball Grid Array (BGA) E. 300 micron balls
soldered connection or a test method. The word ‘‘ should’’ B. Chip Scale Package (CSP) F. 75 micron balls
reflects ‘‘ best processing techniques’’ and is used to reflect
C. Flip Chip (FP) G. 35 mm [1 .38 in]
general industry practices and a suggestion for guidance only.
D. 760 micron balls H. 1 .2x chip size
1
2 APPLICABLE DOCUMENTS
2.1 IPC 1
I PC-T-50 Terms and Definitions for Interconnecting and Packaging Electronic Circuits
I PC-TM -650 Test Methods Manual 2
2. 3 . 1 8 Gel Time, Prepreg Materials
2. 4. 28 Adhesion, S older Mask (Non-Melting Metals)
2. 6. 1 Fungus Resistance of Printed B oard Materials
2. 6. 3 Moisture and Insulation Resistance, Printed B oards
2. 6. 3 . 1 Moisture and Insulation Resistance - S older Mask
2. 6. 3 . 2 Moisture and Insulation Resistance, Flexible B ase Dielectric
2. 6. 1 4. 1 Electrochemical Migration Resistance Test
I PC-92 01 S urface Insulation Resistance Handbook
I PC-771 1 /772 1 Rework, Modification, and Repair of Electronic Assemblies
2.2 American Society for Testing and Materials (ASTM) 3

ASTM D1 50 S tandard Test Methods for AC Loss Characteristics and Permittivity (Dielectric Constant) of S olid Electrical
Insulation
ASTM D2 57 S tandard Test Methods for DC Resistance or Conductance of Insulating Materials
ASTM D792 S tandard Test Methods for Density and S pecific Gravity (Relative Density) of Plastics by Displacement.
ASTM D1 002 S tandard Test Method for Apparent S hear S trength of S ingle-Lap-Joint Adhesively B onded Metal S pecimens
by Tension Loading (Metal-to-Metal)
ASTM D2 471 S tandard Test Method for Gel Time and Peak Exothermic Temperature of Reacting Thermosetting Resins
ASTM D2 556 S tandard Test Method for Apparent Viscosity of Adhesives Having S hear-Rate-Dependent Flow Properties
2.3 Telcordia Technologies, Inc. 4

G R-78-CO RE Generic Requirements for the Physical Design and Manufacture of Telecommunications Products and Equip-
ment
3 TERMS AND DEFINITIONS

All terms and definitions used throughout this handbook are in compliance with IPC-T-50. S pecific terms and definitions,
essential for the discussion of the subj ect, are provided below.
Channeling – Filler settling due to the pitch of the bumps on the underfilled device being too small to allow the filler to
flow through uniformly. The effect can be observed in a scanning acoustic image (for flip-chips) or cross-sectioned sample.
The channeling may also occur due to the presence of flux residue that has reduced the effective gap size.
Board – In this document the term board is used to represent the following:
a. Printed Circuit Assembly (PCA)
b. Flex circuit
c. Ceramic substrate
1. www.ipc.org
2. Current and revised IPC Test Methods are available on the IPC Web site (www.ipc.org/html/testmethods.htm)
3. www.astm.org
4. www.telcordia.com
2
4 BACKGROUND
4.1 Why Is Underfill Needed? Underfilling encapsulants were originally developed to improve reliability of early flip chip
assemblies employing alumina (Al 2 O 3 ) substrates. S older fatigue of the outermost solder bump interconnections caused the
failures in these assemblies. The relatively small thermal expansion mismatch between the silicon die (~3 ppm/°C) and the
substrate (5 - 50 ppm/°C) was responsible for the onset of fatigue during thermal cycling. The temperature range of the
cycling and the size of the solder bump array (closely correlated with the die size) were the main factors determining the
lifetime of the assembly. Inserting a curable liquid epoxy resin between the die and substrate as seen in Figure 4-1 , was
shown to dramatically improve assembly reliability by distributing stress due to CTE mismatch over the entire area of the
die, instead of j ust the solder j oints.
Underfill development accelerated with the introduction of

epoxy-based laminates as substrates for flip chip assemblies
(see Figure 4-1 ). A large CTE mismatch between the laminate A
and silicon (typically 1 2ppm/°C or greater) made underfill a
necessity to prevent very rapid solder fatigue. The underfill
distributes the stresses caused by the CTE mismatch over the
entire area of the die and forces much of the stress into the
substrate. Interfacial delamination between the underfill and
the die, not solder fatigue, is the dominant cause of failure in
underfilled flip chip assemblies.
With the advent of handheld electronic products, use of board

level underfills have become more prevalent. In addition to
imparting thermal cycling reliability, underfills protect chips
B
and interconnections from environmental factors and increase
the mechanical strength of the package attachment. The latter Figure 4-1 Both the Flip-Chip and CSP Underfill in a
is particularly important for chip scale packages (CS Ps), Flip-Chip CSP Soldered to a PCA
which typically pass thermal cycling and environmental tests A. Component Level Underfill
without underfill but fail when subj ected to mechanical shock
B. Board Level Underfill
or impact (e. g. , dropping a cell phone). These forces cause
circuit board flexure that can lead to cracking at the substrate-to-ball interface or at the ball-to-package interface. Underfills
protect CS P solder ball connections by locally stiffening the assembly, thus reducing the strain related to board flexure. In
these cases the CTE is not as critical parameter as Modulus (see S ection 1 2).
When determining the appropriate board level adhesive for product use, material coverage type as well as appropriate mate-
rial properties should be considered. Material coverage will be dictated by necessary enhancement sought for product imple-
mentation and Figure 4-2 illustrates different coverage types.
A B C
G G G
E F F F
IPC-030a-4-2
Figure 4-2 Different Material Coverage Types for Top and X-Sec View (Underfill, Corner Glue, Epoxy Flux)
A. Underfill C. Epoxy Flux E. Cross-Section G. Package
B. Corner Glue D. Top View F. Board
3
Table 4-1 provides some guidelines for coverage selection as well as the appropriate material properties required for
addressing the desired reliability need. In all cases a careful review of the physical properties and product testing with board
level adhesives would need to be conducted to confirm the reliability enhancement has been achieved.
Tabl e 4-1 G u i d el i n es for Eval u ati on
R el evan t P h ysi cal
Fu l l U n d erfi l l E d g e/Corn er B on d Properti es of Ad h esi ve to
R el i abi l i ty N eed Coverag e or G l u e Coverag e E poxy Fl u x Coverag e Ad d ress R el i abi l i ty N eed
Not typical; only a Not typical; limited

Thermal Cycling Yes localized benefit benefit due to low Tg , CTE, Modulus
can be achieved material coverage
Drop or Shock Yes Yes Yes Modulus, Adhesion
Bend or Twist Yes Yes Yes Modulus, Adhesion
Fluid/Gas Protection Yes No No Adhesion, Chemical
Resistance
Vibration Yes Yes No Modulus, Adhesion
4.2 Types of Underfill Underfills are typically two-part pre-

mixed and frozen epoxy systems (see Figure 4-3 ). These O O O
C C OH
epoxy systems may contain inorganic particles to modify the O R + CH 2 CH O R
C OH C O CH 2 CH
physical properties of the system. These inorganic particles
O O
are called fillers. S ee 6. 1 for more information on filler sys-
tems.
O C(CH 3) 2 OCH 2 CH CH 2
O
CO
4.2.1 Capillary Underfill These low viscosity liquids are R
designed to flow under a component by capillary action, wet- CO
ting to the component and substrate surfaces and encapsulat- O
ing the solder j oints. Underfills designed for assembly-level
O C(CH 3) 2 OCH 2 CH CH 2
IPC-030a-4-3
flip chips generally cure in five minutes or less to form a hard
Fi g u re 4-3 E xam pl e of Epoxy Ch em i cal Reacti on
seal with high adhesion to both the component (flip chip or
CS P) and the board - those types of underfills are typically
called snap-cure underfills. Faulty or skewed components
must be detected prior to underfill cure, since capillary flow
underfills are typically permanent and even in the case of
reworkable underfills reworking a part with adhesive has yield
and time impacts.
Underfill is applied close to the edge of a flip chip or CS P to

enable capillary forces to encapsulate the gap between the
component and the board. Dispensing of capillary underfill
materials requires specialized equipment to achieve the accu-
racy and precision required for high volume assembly (see
Figure 4-4). At minimum, the dispenser must reproducibly
position successive assemblies and apply a pre-determined
volume or weight of underfill to the edge of the component.
Heating the substrate is a common secondary requirement that
accelerates capillary flow, and minimizes voiding. Cure is usu-
ally accomplished in belt-style reflow or batch curing ovens.
4.2.2 Fluxing (No-Flow) Encapsulant Fluxing encapsulant

can be applied in different ways. Two common methods are
Fi g u re 4-4 I m ag e of N eed l e Di spen si n g of an U n d erfi l l
dispensing directly onto the board prior to component place- (B ottom Si d e Vi ew)
ment or the component can be dipped into the fluxing mate-
rial and then placed onto the board. A fluxing function incorporated into the underfill combines the component attachment,
soldering, and underfill cure processes thus improving throughput and cost by decreasing the number of manufacturing steps
and decreasing cost by a reduction in floor space.
4
Fluxing underfill is applied directly to the attachment site of

the board prior to component placement. The flip chip B GA
(FCB GA) or CS P is then placed into the underfill (see Figure
4-5). During reflow, the underfill acts as the flux, enabling
interconnect formation and self-centering prior to curing of
the underfill. Ideally, underfill cure is completed in the reflow
oven. Otherwise, subsequent reflow or deliberate heat treat-
ment is required to complete underfill cure.
Fluxing underfills are different from capillary flow materials

in several ways. Viscosities are typically much higher than
that of capillary underfills. The inorganic fillers used to
modify the properties of capillary underfill systems cannot be
used in no-flow systems since the inorganic filler would inter-
IPC-030a-4-5
fere with the electrical connection between the bumps and the
pads. Therefore, no-flow systems typically have higher CTE Figure 4-5 Fluxing Underfill Process
than capillary underfill systems.
After the underfill and reflow processes, the additives that provide the fluxing function are inert and should not compromise
electrical reliability with proper process conditions. If there is concern for a specific formulation and processing condition,
testing can be performed using industry standard S IR/ECM methods, both IPC (IPC-TM-650, TM 2. 6. 3 , S urface Insulation
Resistance, Fluxes) and Telcordia (GR-78-CORE, Generic Requirements for the Physical Design and Manufacture of Tele-
communications Products and Equipment).
4.2.3 Removable/Reworkable Underfill Occasionally, malfunctioning FCB GA or CS P components are discovered in the
factory or in the field. Conventional underfills are not reworkable once cured without destroying the substrate and/or pads,
requiring that a malfunctioning assembly be discarded, sometimes at considerable cost. The use of reworkable/removable
underfill enables repair of the assembly after underfill dispense and cure.
Reworkable underfill is typically dispensed and processed similar to other underfills (capillary or fluxing). Different formu-
lation characteristics enable them to soften or degrade when sufficiently heated. The assembly repair process is similar to
that of other solder surface mount technology (S MT) components with the exception of extra steps or extra time required
to remove residual underfill before and after component removal.
4.2.4 Cornerbond Adhesive (Dispensed or Placed Prior to Reflow Process) Another class of adhesives used for enhanc-
ing reliability is placed only along the edges of the component and can be cured during the reflow process. Due to the close
proximity of the printed solder paste and the edge of the component, this process has limited use and is dependent on prod-
uct layout design as well as placement accuracy for both the part and the adhesive. The implementation challenges for pre-
reflow glue have led to other adhesives that are applied and cured after the S MT process.
Figure 4-6 is an example of the process used for reflow cured cornerbond underfills. The cornerbond material must not cure
until the solder paste and solder ball wet and j oint formation has completed.
4.2.5 Corner Glue/Corner Tack Underfills (Applied After SMT Reflow) Very similar to the corner bond materials, this
class of adhesive is only applied along the perimeter of the component needing reliability enhancement. The material is
applied after the S MT reflow process and can be cured via a thermal process in an oven or using UV light.
Figure 4-7 illustrates a thermally cured corner glue/corner tack on the left and on the right a UV-cured corner glue/corner
tack material.
5
A B D
IPC-030a-4-6
Figure 4-6 Reflow Cured Corner Bond Process

A. Printing solder paste and dispensing cornerbond C. Reflow
B. Mounting D. Melting solder, bottoming CSP or curing CB
Figure 4-7 Thermally and UV Cured Corner Tacks
5 MECHANICAL CONSIDERATIONS
5.1 Footprint Design The drive for area array technology often stems from a need to miniaturize, to obtain electrical per-
formance improvements, or to accommodate interconnect density demands. All of these needs highlight the critical nature
of a comprehensive design - a footprint design that is completed in concert with the design of the next level of assembly.
One common design mistake is to place passive or other components too close to the device requiring underfill. Adj acent
components can make it difficult to get the dispensing needle into the proper position and often create a surface area that
causes the underfill material to wick into areas where it is not needed or wanted. A dead area of 3 X the dispense needle
inner diameter is suggested as a keep out area along each edge. S ome newer dispensing technologies (e. g. , j etting) can
reduce this keep-out area. However, this area can never be zero since the underfill material itself will flow out during dis-
pense and encapsulate adj acent devices. In addition to designing keep out zones for adj acent components, test points also
need to be clear of the underfill flow out area so that they are not encapsulated leading to problems with in-circuit testing
(ICT).
Furthermore, the masking under the package should be uniform. S pecifically, uneven surfaces and unmasked areas allow for
uneven flow out and entrapment of air bubbles under the package. Also, in some cases trench-style solder mask definition
can cause difficulty in underfilling.
6
The bump/pad spacing, gap height, and pitch all are important when selecting underfill. The underfill formulation selected
needs to be compatible with the product design parameters.
The placement of un-covered vias under and near the underfilled device can be detrimental to the performance of the under-
fill, by inducing voids under the device or thieving material from the fillet. The surface tension of the underfill will deter-
mine the maximum open via size that can be used. Tenting vias will insure that material will not flow into vias, and elimi-
nate the risk of voiding and reduction of fillet size.
Also, when there are open vias in the underfill wet out area, the heater platen or hot-plate will pull underfill through the
board through-hole due to capillary forces. In order to minimize this effect, the board should be elevated off the heater plate
or other mitigation strategies must be utilized to prevent underfill flow through board through-holes.
5.2 Gap Size Gap is the space between the assembled package and board with the bump, or j oining structure serving as
the standoff. The appropriate filler size and dimension is based on the gap size of the device to be underfilled. A commonly
1
employed guideline is to choose a filler that is ⁄3 the size of the smallest gap. B ump pitch in addition to gap size can also
be a determining factor for filler size selection, if it is smaller than the gap size. Gap size should be measured as a function
of actual process parameters and not j ust estimated from the bump size since the gap height is affected by the solder paste
deposit volume and the bump collapse during reflow. Also, the presence of flux residues can reduce the effective gap size
to a fraction of the calculated value. Excessive eutectic solder bump collapse due to overly large bonding pads and can also
reduce the gap to a level where underfill will not flow. The reflow profile can also change the gap size so filler size selec-
tion should include product process parameters in addition to design parameters.
5.3 Pad Redistribution A bumping alternative for those chips which weren’ t designed specifically for flip chip assembly
is to redistribute the pads into a pad configuration which meets the bumping design rules and which can be flip chip
assembled more readily. These redistributed pads should meet all the design criteria presented in this section. When redis-
tributing the bumps, the pattern on the device should be such that the forces on the die during reflow are balanced.
6 UNCURED UNDERFILL CHARACTERISTICS

Underfill is a liquid prepolymer system that can be filled or unfilled. In the case of filled systems, the primary use of the
filler is used to lower the CTE of the final cured system. Understanding the raw underfill material characteristics is impor-
tant for determining the best methods and procedures for dispensing and processing.
6.1 Filler Properties The two most common fillers used in underfill systems are silica (S iO 2) and alumina (Al 2 O 3 ). How-
ever, alternatives such as boron nitride (B N), zinc oxide (ZnO), and other materials may be used (see 6. 1 . 2). The type of
filler, particle size, particle distribution, particle geometry, particle surface area, density, and surface treatments can be
important criteria in the filler selection process and will depend on application needs. The rate of filler settling is dependent
upon the density of the filler, the particle size of the filler, the specific gravity of the resin(s), and the viscosity of the filled
encapsulant. S ee 8. 2. 2. 2 through 8. 2. 2. 4 on phase separations for further information on filler settling. The size and mor-
phology of the filler particles in underfill materials is controlled. The supplier shall provide some indication of a process
maturity, stability and capability for the control of particle sizes upon request.
6.1 .1 Filler Size S uccessful underfilling requires that the filler particles are smaller than the effective gap size. The smaller
the actual gap size the more critical the selection of maximum filler particle size. If the filler diameter is half the gap height
or greater, empirical tests have shown instances of highly restricted flow. As such, the maximum filler diameter should be
1
⁄3 the size of the smallest gap through which the material is expected to flow (e. g. , package-to-board and j oint-to-j oint).
Filler particles may also become trapped between bumps if the maximum particle size exceeds the effective bump pitch.
Flux residue often surrounds bumps and can significantly reduce the effective bump pitch. Considering criticality of particle
sizes, demonstrating statistical control that ensures process capability with predictable variability around nominal targets is
strongly recommended.
The supplier should report the maximum particle size of the filler to determine compatibility with the user’s minimum gap
size. Particle size of the filler can be determined by using progressively smaller mesh screens or by means of a particle size
analyzer. Another method is to cross-section the underfill, polish it and then examine the cross-section under a microscope
and measure the particles in the field of view. This method will give a rough estimate of filler size distribution and relative
size of filler to relevant gaps in the surface mount assembly.
On the other hand, very small filler particles may reduce flow rate, depending on their density. Therefore, a balance of par-
ticle size distribution is needed to provide an underfill that does not become trapped but also does not significantly impact
the flow of the material.
7
6.1 .2 Filler Material Type There are a number of different types of fillers that can be used in underfill encapsulants. While
silica is most prevalent filler types, in high power dissipation applications a more thermally conductive filler such as alu-
mina or boron nitride may be used. The underfill supplier should report the filler type to users and filler type should be
selected to meet application needs.
6.1 .3 Filler Content The filler loading of an underfill has a significant impact on the material’s capacity to alleviate CTE
mismatch. For materials designed to alleviate CTE mismatch, filler loading should be tailored with the product temperature
cycling requirements while balancing manufacturing targets. Underfill suppliers prepare formulations in terms of weight
percent versus volume percent and, depending on the filler type, the weight percentage may be significantly different than
the volume percentage. However, board level underfills can be formulated with no filler loading depending on product needs
for reliability and manufacturing.
6.1 .4 Density Typically, suppliers report density of the underfill and the density will be determined by the base resin, filler
type, and the filler loading of the formulation. Typically, AS TM D792 or similar method is used to determine density. In
low-viscosity underfill systems, filler settling may be a concern. S ee 8. 2. 2. 2 through 8. 2. 2. 4 for more information. Underfills
with higher density fillers may be more susceptible to filler settling or stratification particularly during high-temperature
cures.
Thermal gravimetric analysis (TGA) residue of a ‘‘ wet’’ specimen may be used to determine relative filler content. The TGA
may also be used to determine if filler settling may have occurred by curing a highly vertical but thin specimen and ana-
lyzing the top and bottom of the specimen for relative filler content. S everal AS TM methods (www. AS TM. org) can be used
to determine filler content using TGA.
6.2 Prepolymer Properties
6.2.1 Viscosity Although viscosity is reported in a material data sheet for a standard condition, viscosity will likely change
over the life of the material. Viscosity is important to flow rate and dispensability of the material. The supplier material data
sheet should be referenced with respect to the method and definition. An example viscosity test method is AS TM D2556 -
Test Method for Apparent Viscosity of Adhesives having S hear-Rate-Dependent Flow Properties.
6.2.2 Gel Time The gel time refers to the point where the liquid begins to exhibit pseudoelastic properties (becomes vis-
cous and tacky). This gel time will vary, depending on the curing system used and any preheating of the encapsulant prior
to or during dispensing.
After the underfill process is complete, users may elect to gel the underfill so that many devices can be collected and final
cured in a batch mode. The gel process may reduce issues associated with filler settling and moisture absorption that can
occur if underfill is maintained under the device uncured for extended periods of time. Precautions need to be taken to ensure
the gel point is not attained while the underfill encapsulant is still in the syringe or during the underfill process due to expo-
sure to the heated stage.
Underfill suppliers should provide gel time requirements for their respective encapsulants, per either of the following two
methods: AS TM D2471 S tandard Test Method for Gel Time and Peak Exothermic Temperature of Reacting Thermosetting
Resins or IPC-TM-650, TM 2. 3 . 1 8, Gel Time, Prepreg Materials. A test performed on a hot plate known as the stroke gel
test can be used to determine gel times in a cursory manner. Typically, the gel time for underfills is dependent on time and
temperature. Typically underfill materials are dispensed at elevated temperatures as this will allow the material to wet and
flow out more quickly, increasing through put rates (e. g. , units/hour). However at the 70-90 °C [1 58°-1 94 °F] temperature
range, the underfill may begin to polymerize or gel depending on the formulation. It is important that the underfill process
conditions are optimized to ensure that gel onset does not occur during underfilling of the device.
6.3 Material Compatibility
6.3.1 Flux Compatibility Consideration should be given to underfill compatibly with the solder paste or tacky flux residue
produced during the soldering process. This especially applies to ‘‘ no-clean’’ or low residue fluxes that may leave a residue
on the board or on the underside of the package. In addition, the more aggressive fluxes used for Pb-free soldering tend to
have a greater impact on underfill compatibility.
The no-clean flux residues can interfere with the underfill by two different mechanisms. First the physical presence of the
flux residue can impede or prevent the underfill from wetting the bumps and/or the package (see Figure 6-1 ). S econd, the
8
acid groups in the flux residue can react with portions of the
epoxy chemistry, essentially acting as an unintended catalyst
or cure inhibitor. Flux residue and underfill chemical interac-
tion can alter the final underfill properties leading to new fail-
ure modes.
For some applications flux residues are cleaned prior to the

underfill process. If cleaning is conducted, care should be
taken to ensure the area under the device is dry. As residual
moisture or solvent can lead to underfill voids or at worst cure
inhibition. However, trace amounts of residue may remain
from the flux or the cleaning process and compatibility with
underfill materials should be verified.
The impact to adhesion strength of underfill with relevant

interfaces such as solder mask is particularly important to
evaluate in the presence of residue. Adhesion differences may Fi g u re 6-1 SE M I m ag e Sh owi n g Poor U n d erfi l l Ad h esi on
be exacerbated with change in temperature and humidity con- to a B u m p Du e to Fl u x R esi d u e
ditions. Parts should be examined for voiding that may result

from the presence of flux residue or interaction effects between underfill and the flux residue. The amount, and even the
composition, of flux residues can vary with the reflow process shifts, so good process control and monitoring will play a
significant role in determining flux compatibility.
6.3.2 Board Surface Compatibility The maj or interaction between the underfill and the board surface occurs with the sol-
der mask. In general, adhesives have better adhesion to rough surfaces, therefore a matte surface, rather than a glossy sur-
face, provides a better surface for mechanical adhesion. Two common modes of failure between the underfill and board
occur due to:
• Moisture
• Under-cured solder mask
The board solder mask can be a source of absorbed moisture. A capillary underfill should be applied to a dry surface. In the
case of a no-flow underfill system, absorbed moisture that outgases during the reflow process can cause voiding in the
underfill leading to lower reliability.
Another common solder mask related failure mode is incompletely cured solder mask. Incompletely cured solder mask will
absorb more volatile materials during surface mount processes, and these chemicals can then outgas during the underfill
heating steps leading to voids and delamination.
In cases with no solder mask, adhesion to laminates such as FR-4 is typically very good. Nonetheless it is advised, that
compatibility between the underfill and the relevant board surface should be tested through an adhesion metrology.
6.3.3 Component Surface Compatibility There are a variety of component surfaces to which underfills need to adhere.
Good adhesion to these surfaces is critical in order to eliminate delamination. Also, j ust as moisture can evolve from the
solder mask into the underfill (see 6. 3 . 2) moisture can be absorbed into these materials as well and then exit into the under-
fill, if the moisture is not driven from the material prior to underfilling.
6.4 Alpha Particle Emissions Certain levels of alpha emissions (from uranium and thorium isotopes) are known to cause
soft errors in memory devices (i. e. , DRAMs and S RAMs). The permissible levels are very application specific. In general,
the proximity of the alpha emission source to the bump diameter and chip structure needs to be considered to estimate the
soft error rate. In order for alpha particles to cause soft errors, the following conditions must be met:
• Alpha particles must be located near the active circuitry area,
• Alpha particles must have a direct line of sight to the active circuitry area, and
• Devices need to be sensitive to alpha radiation.
The most prevalent sources of alpha radiation in a typical flip chip assembly are; high lead solder (0. 05 α/cm 2
-h -
1 0. 0 α/cm -h), 2
alumina (0. 1 α/cm 2
-h), chip underfill (0. 002 α/cm 2
-h - 0. 02 α/cm 2
-h) and plastic mold compound (0. 04
α/cm 2
-h). In terms of underfill materials, the filler material is the primary source of α-emissions. In comparison to the high
9
lead solder, the underfill encapsulant is less of a concern for α-emission. Underfill suppliers typically use filler material that
has very low α-emissions (in the ppb range). Presently, particle barriers; (i. e. , passivation layers) and optimized circuit
designs are used to minimize exposure of sensitive circuitry to α particles. S ynthetic silica fillers and some naturally occur-
ring fillers are available with nearly zero emissions, but at higher cost.
7 MATERIALS PACKAGING, HANDLING, AND STORAGE
7.1 Packaging and Containers Underfill is most commonly packaged in sealed plastic syringe bodies designed for auto-
matic dispensing machines. The syringes are available in a wide variety of sizes based on metric volume in cubic centime-
ters (cc). Most suppliers place a human or machine-readable label directly on the syringe body that indicates product, lot,
expiration data, and other pertinent information. The human readable label for very small syringes, such as 3 cc, may be
placed on the external bag however most production lines use 1 0 cc or larger syringes that can accommodate the label. S up-
plier provided labeling should be capable of surviving designated storage conditions.
Common syringe sizes include 1 0 cc and 3 0cc syringes and 6 oz cartridges. It is typical to select syringes or cartridges that
match the material’s pot life to the volume of production units to minimize cleaning and syringe change over.
Note: Many sizes are available and the user should select a size that will fit the user’s specific equipment.
The syringes should be fitted tightly with end caps and seals to prevent leakage during shipment and warm up. S everal
plunger styles are available. The plunger should permit smooth continuous flow of the material. S yringes and closures should
be selected that are compatible with the selected dispensing equipment. Optionally, the syringes can be sealed in plastic bags
to prevent moisture condensation upon removal from freezer storage.
7.1 .1 Voids/Bubbles in Packed Material Containers should be maintained so that they are free of air bubbles. Air bubbles
can interfere with dispensing and cause voids under components. Air bubbles that cannot be removed from the fluid path
can result in voids in the dispensed fluid. Ideally, these air pockets can be purged into waste or in areas that do not affect
product performance. However, they can often create a void under the chip that can become initiation points for cracking
and delamination. Voids can be introduced in the syringe if the syringe undergoes a temperature shock, such as being
removed from the freezer or from a dry ice filled container. To avoid freeze-thaw voids, the user should take precautions so
that syringes stored at -40 °C [-40 °F] or lower do not undergo a rapid thermal shock or repeated freezing and thawing. An
example of a procedure that can reduce this phenomenon is:
1 . Upon receiving material in Dry Ice (-80 °C [-1 1 2 °F] ), remove from shipping box, do not open inner box, put into freezer
(-40 °C [-40 °F] ) for a minimum of 8 hours before removing to thaw material. Do not open inner box for any reason,
extreme temperature difference can cause air voids.
2. When taking material out for use, remove whole inner box, do not open, let stand at room temperature (RT) for 1 hour
(any size syringe) in the box before opening.
3 . After 1 hour, open inner box and remove needed syringes/cartridges. Place unwanted syringes back in freezer in inner
box, and mark box ‘‘ partial’’ with how many are inside for future reference. Material will not be completely thawed;
hence preventing the syringe expansion that could allow air into the material. The slight thawing of the material will not
affect the pot life of material going back into the freezer.
4. Continue to thaw syringes at RT with tip down until the syringe equilibrates with RT. Vertically thawing material will
minimize air in the fluid path.
7.2 Storage Conditions Underfill is commonly a two-part epoxy that is pre-mixed and frozen. It usually needs to be stored
below room temperature and often materials with high filler loads (above 40%) need to be stored in a low-temp freezer (e. g. ,
-40 °C [-40 °F] ) in order to have an acceptable shelf life. Cold storage prevents filler settling by greatly increasing viscos-
ity or actually solidifying the underfill and more importantly, slows the curing process, thereby providing an acceptable shelf
life. Other unfilled materials commonly used for CS Ps can be stored at -20 °C or at 5 °C [-4 °F or 41 °F] for up to 6 months
or a year. Follow supplier recommendations for optimal storage conditions and acceptable shelf life to ensure the material
remains useable in terms of viscosity, flow, and other performance criteria.
Newer formulations are now available that do not need to be stored in a low temp (e. g. , -40 °C [-40 °F] ) freezer, but in a
standard freezer or refrigerator. These newer chemistries are designed such that the degree of cure will be slowed so the
material is stable at -5 °C to 8 °C [23 °F to 46 °F] rather than -40 °C [-40 °F] . Their increased temperature stability also
results in an extended pot life at ambient temperatures.
10
7.3 Preconditioning Underfill that is freezer-stored must be brought to ambient conditions prior to dispensing. Conden-
sation of water from the atmosphere will occur with cold samples and should be wiped off the container prior to use. S ome
products may be packaged with a sealed plastic bag surrounding the container. This outer covering should be removed only
after the product is at ambient temperature in order to prevent moisture contamination. One of the most common causes of
voids within a syringe is improper handling and care should be taken to minimize any thermal shock to a syringe. For
example, syringes shipped in dry ice (-80 °C [-1 1 2 °F] ) should not be removed directly from the dry ice to ambient. Often
times when this occurs a ‘‘ cracking’’ sound can be heard, and this noise is due to a void forming in the syringe from CTE
mismatch between the syringe material and the frozen underfill. The proper protocol for thawing syringes is to remove
syringes from dry ice and place syringes into (for -40 °C [-40 °F] storage materials) the low-temp freezer. After the mate-
rial has stabilized in the low-temp freezer thaw material as needed at room temperature to minimize voiding in the syringe.
7.4 Pot Life Underfill remains useable only when it can be dispensed, flowed, and cured properly. Most underfill products
change over time, even under freezer conditions and slowly polymerize. This polymerization produces an increase in vis-
cosity and thus reduces dispensing rate and underfill flow rate under the component. The following tests are intended to
measure changes caused by filler settling and polymerization as well as reduction of polymer catalytic activity. The life of
the material can be measured in a variety of ways (for example a 50% change in viscosity). Dispensing should be done
without significant increase in throughput rate through the needle or in the flow out time and significant changes are an indi-
cation the material’s useful life is over or nearing its end.
7.4.1 Viscosity Change Filler settling and polymerization will result in a viscosity increase. Underfill should be measured
after storage using the method in 6. 2. 1 and an increase of 20% may be acceptable for most applications. However an
increase of 50% or more is considered unacceptable since dispenser control may become difficult and underfill flow rates
may also decrease.
7.4.2 Flow Rate Change A test should be developed by the user to be run with dispensing equipment. The dispense flow
rate measurement method in 8. 4. 1 can be used to assess the change. The flow rate of the underfill is monitored as a func-
tion of time. Thus the time per amount dispensed is typically recorded over an 8-hour period. Flow rate change is applica-
tion specific and tolerances should be determined for each process. This is a critical parameter for capillary flow underfill
materials (see the glass slide test referenced in 8. 2. 2. 1 ).
7.4.3 Settling Test A simple gauge of filler settling is to measure the first and last dispensed portions from a needle
syringe that has been stored vertically. If settling occurred, the first sample will contain more filler and have a higher vis-
cosity and the last portion of underfill dispensed would consequently have less filler and lower viscosity. Use the method in
6. 2. 1 for viscosity measurements. If settling is visible within the package prior to dispensing then the material is not recom-
mended for use.
8 APPLICATION PROCESS
The material should only be used at the supplier recommended temperature which is typically at ambient. A syringe-cooling
device is sometimes necessary to keep the syringe at ambient conditions when inside an automatic dispenser that is heating
the board. A plastic syringe, which is the most common container, may be used manually by removing the dispense end cap
and installing a needle of the appropriate size either recommended by the material supplier or determined through process
characterization. A plunger may also be necessary for manual dispense applications as typical packaging is designed for
automatic dispensers. Alternately, the pressure supply end may be fitted with a cap attached to a pneumatic dispenser. Pro-
duction assemblies typically utilize an automatic dispenser designed to accept standard syringes. The working temperatures
of underfill materials will vary by location in the dispenser. The syringes are held at ambient for the machine, but in hos-
tile environments, ambient conditions can be in the 40 °C [1 04 °F] range. The material is dispensed through a needle, which
is often heated to 3 5-50 °C [95-1 22 °F] . The work piece is pre-heated, heated during the dispense sequence, and heated after
the dispense to promote underfill flow. Typical dispense and post dispense temperatures are in the 65-80 °C [1 49-1 76 °F]
range. However, this temperature should adhere to supplier recommendations as some underfills start their curing process at
temperatures as low as 70-1 00 °C [1 58°-21 2 °F] .
While the package is positioned under the dispensing head, the PCA should be heated to increase the flow rate of underfill.
B oth the dispensing needle and the PCA platform may be heated. Dispensing equipment is now available specifically for
underfill and material suppliers should specify the recommended temperature range for optimum flow; between T1 (tempera-
ture where viscosity is significantly lower) and T2 (gel point). When using no-flow or fluxing underfills, the material is dis-
pensed onto the board and the package is placed into the dot of material j ust prior to reflow.
11
8.1 Pre-Application Board Preparation A dry PCA is required to prevent moisture related voiding of the underfill during
reflow and cure. B oards exposed to ambient air for more than four hours should be thoroughly dried before underfill
dispense. Typically, one hour at 1 25 °C [257 °F] is generally enough to dry laminate substrates up to 500 µm thick, but two
hours or more may be required to dry thicker boards. In cases where off-line drying is impractical, shorter exposures to
reflow conditions may provide sufficient drying. However, the moisture absorption due to sit-times at ambient conditions
during production should be carefully monitored to minimize the incidence of underfill voiding. This drying is particularly
critical in the case of no-flow underfill application where the boards have not gone through a reflow process prior to the
application of the underfill.
8.2 Application of Capillary Flow Underfill The first step in the capillary underfill process is to attach the device to be
underfilled. As such, it is necessary to flux the solder bumps and pads to promote solderability and there are 3 ways to attach
the package to the board:
a. Printing (flux and/or paste)
b. Dipping (the package into a doctor bladed flux film) (see Figure 8-1 )
c. Jetting flux onto the pads
D
IPC-030a-8-1
Figure 8-1 Fluxing Underfill

A. Moving flux pool plate
B. Flipping squeegee home position, flux layer 250 µm
C. Dipping position, flux layer ‘‘x’ µm (e.g., 1 30-1 49 µm)
D. Placed
12
The amount of flux residue should be minimized. This is typically accomplished by:
a. Reducing the solids through using a thin flux or a formulation that fully or mostly volatilizes
b. Minimizing the total amount used
c. Ensuring compatibility with the underfill formulation
d. Cleaning the flux residue after reflow and before underfill
For B GAs, the residue from the solder paste can also narrow the effective gap size for underfill flow. The impact on gap size
will be more critical for fine pitch products, but for loose pitch products the gap size is typically significantly larger than the
filler particle size and is not as critical.
8.2.1 Dispensing Procedures Underfill must be dispensed close to the edge of the assembled package. A rule of thumb
1
is ⁄2 the outside diameter of the needle is the correct gap between needle edge and die edge.
The fill stage involves dispensing underfill at one or more edges of the package and allowing the material to flow to oppo-
site sides by capillary action. It is important that a reservoir of liquid be maintained at the edge of the device so that air
bubbles are not introduced. Once underfill has flowed completely and is present at all sides of the device, the fillet step is
carried out.
The fillet step involves dispensing additional underfill along
A
the device edges not used in the fill step. A fillet on all sides
of the die is thought to increase reliability and to reduce the
possibility of delamination or die cracking. B ased on the
underfill chemistry selected, an adequate fillet may be formed
during the fill process, therefore a filleting step may not be
B
required. Cosmetics are also improved with a fillet step and
this can be important in packages such as flip chip-B GA. Fig- C
ure 8-2 shows underfill dispensing.
8.2.1 .1 Dispensing Patterns S everal dispensing patterns B

are used for applying underfill to packages. Underfill can be IPC-030a-8-2
dispensed from 1 , 2, or 3 sides, but at least one side must be
Figure 8-2 Underfill Dispensing
left open so that displaced air will not be trapped under the
A. Fill step
device. The simplest pattern involves applying underfill to a B. Board
single side and maintaining a reservoir as material wicks C. Fillet for maximum reliability
between the package and board. S ome of the key differences
between these dispense patterns are speed, fillet width, and
risk for air entrapment. ‘‘ I’’ passes (one side dispense) are C D F
used to minimize the fillet region, and multiple passes may be
necessary to dispense the volume of material desired. ‘‘ L’’ (A) E
passes are quicker, and require fewer passes to achieve the
same volume. However, the flow front under the package
should be monitored for air entrapment and the fillet width
G
should be considered with regards to the proximity of other
devices. ‘‘ U’’ passes are rarely used as the speed increases are
(B) H
not significant, and the risk of air entrapment increases. When
flow is complete, if the material has not formed a fillet on all IPC-030a-8-2
sides, underfill can be applied to the remaining sides and in Figure 8-3 Examples of Dispensing Patterns (The Fillet is
the case of an ‘‘ I’’ pass dispense a ‘‘ U’’ pattern would be used Shown in a Different Color for Clarity Only)
to form a fillet. The pattern, called line + ‘‘ U,’’ is shown in A. Line + ‘‘U’’ pattern E. +’’U’’
Figure 8-3 (A) and is recommended for larger and more com- B. ‘‘L’’ pattern F. Fillet
plex products since it has the lowest chance of air entrapment. C. Line G. ‘‘L’’
For smaller devices when the fast-flow and self-filleting mate-
D. Flow H. + ‘‘L’’
rials are used, the second ‘‘ U’’ pass is not required. The mate-
rial can flow under the device, exit the opposite side and form a fillet. This flow-out and self-filleting can occur during a
wait step in the dispense process or can occur during the ramp-up of the cure cycle, thereby increasing throughput. Another
common flow pattern recommended for fast processing of larger devices involves dispensing an ‘‘ L’’ of material as shown
in Figure 8-3 (B ).
13
Typically, for fluxing underfills, a single large dot is placed at the center of the array pattern. There may be variations on
this pattern if the component has an odd form (elongated or non-symmetric bump pattern) or is a large size.
8.2.1 .2 Process Parameters All of these dispense patterns

can be achieved using a manual needle dispense, automated
needle dispense, or j etting (see Figure 8-4). Typically j etting
has higher throughput, can accommodate tighter keep out
zones, and has tighter tolerances for material dispense weight.
However j etting may not be compatible with all chemistries as A
it requires droplets to individually break to achieve high
repeatability. Depending on material viscosity and thixotropic
IPC-030a-8-4
behavior most underfills can be j etted, but other materials
such as corner glue may require needle dispense. As such, the Fi g u re 8-4 I m ag e Sh owi n g B oth N eed l e Di spen si n g an d
selection of needle dispense versus j etting should be discussed J etti n g for U n d erfi l l i n g at th e Ch i p Level
between both the equipment and material suppliers to obtain
recommended dispense methods and preliminary dispense
parameters.
Accuracy of the amount of underfill fluid dispensed, and the tolerances of the underfill gap will influence the varia-
tion in fillet shape, height, and width. Changes in fluid viscosity over the working life of underfill fluids can change the
flow rate of the fluid for time, pressure, and auger pump dispensers. Periodic recalibration, either manually or in an auto-
mated fashion, can prevent changes in amount of fluid dispensed. S ervo-controlled piston or linear pumps that are insensi-
tive to viscosity changes can also be used to control dispensing accuracy. Multiple pass dispensing patterns can help reduce
or control the fillet width. The timing of the fillet pass can be critical in some cases to prevent air from being introduced
on the dispense side and to ensure proper flow fronts that minimize the incidence of underfill voiding.
8.2.2 Application Problems

A
8.2.2.1 Air Entrapment Figure 8-5 shows the air entrap-
ment phenomenon caused when material is dispensed from
more than one side of a complex device. Flow tends to pro-
ceed most rapidly along the perimeter bumps.
Although imaging methods or the use of a transparent glass slower flow

die can identify trapped air, a simpler method is to clamp the
bumped die or package to a glass slide and observe the flow
pattern. Alternatively, the bumps can be bonded to glass with
adhesive as described in 8. 6. 2.
A IPC-030a-8-5
There are 3 primary types of void formation. 1 ) Entrapped air
in the syringe flows into the gap and creates a void. 2) The Fi g u re 8-5 Ai r E n trapm en t (U n d erfi l l was Di spen sed on
flow out and dispense pattern allow for the material to flow th e Left E d g e of th e Packag e)
A. Fast fl ow
around a pocket of air and entrap it. This is most common in
complex high I/O patterns (large die, small gap) or in devices
where the s o lder mas k and b all layo ut is non- unifo rm
(trenches exist) 3 . ) Chemical reactions with flux or other
A
materials induce out gassing (moisture and flux residue are the
most common sources). This glass slide method is best used
as a vehicle for approximate or comparative results/studies,
i. e. , ‘‘ benchmarking,’’ but may not give exact results as the
glass surface may not replicate the properties of the actual B
surface and topology.
IPC-030a-8-6
8.2.2.2 Gravitational Phase Separation Gravi tati o nal
Fi g u re 8-6 G ravi tati on al Ph ase Separati on
phase separation may occur when filler settles out in low-
A. Packag e
viscosity underfills, and is exacerbated if they are left at room
B. B oard
temperature for long periods (see Figure 8-6). Heating that
14
can substantially reduce viscosity and can further exacerbate

the settling problem. The result will be a greater concentration
of filler near the board surface and resin enrichment at the A
substrate interface. Long periods between dispense and cure
or very slow ramp up to cure temperature can result in exces-
sive settling, although a small amount will not significantly
affect reliability. If filler settlement cannot be resolved with
the material formulation, one manufacturing solution is to gel B
the underfill soon after flow out is completed.
IPC-030a-8-7
8.2.2.3 Dynamic Phase Separation Dynamic phase separa- Figure 8-7 Dynamic Phase Separation
tion may occur when the organic phase moves rapidly under a A. Package
device and past bumps by capillary (surface tension) action B. Board
(see Figure 8-7). The denser filler will tend to move slower
than the liquid phase so that filler is concentrated at the dis-
pense end of the device. S ome separation is normal for fast
flow underfills. Cross-sectioning will detect both types of
A
phase separation shown in Figure 8-8.
8.2.2.4 Filtering Phase Separation A phenomenon, called B

‘‘ filtering,’’ can cause drastic and unacceptable phase separa-
tion. This happens when filler particles become trapped
C
between the package and board or between bumps (see Figure
8 -8 ). Filtering phase separation is easily seen in a cross- IPC-030a-8-8
section. In extreme cases, no filler particle or only the small- Figure 8-8 Filtering Phase Separation
est ones, are seen beyond the blockage point that is usually a A. Package
bump. The problem is seen in assemblies with low bump B. Board
height (<50 µm) and fine pitch (<250 µm). Although engineer- C. Restriction
ing calculations may conclude that both gap height and pitch
are sufficient to allow the unhindered passage of filler particles in a particular underfill, presence of contamination or flux
residue can significantly reduce both. It may be necessary to cross-section in both X and Y directions to visualize the prob-
lem. S ee troubleshooting described in Section 1 4.
8.3 Application of No-Flow/Fluxing Underfill No-flow/Fluxing underfills are typically unfilled and of lower viscosity than
conventional underfills and encapsulants. Application by brushing or by dipping the chip is not recommended unless expo-
sure of the material to atmospheric moisture can be prevented.
8.3.1 Dispensing Pattern A simple dot-shaped dispense pattern should be acceptable for small packages but for larger
packages, an X-shaped pattern may ensure more material at the corners. For those substrates having an aspect ratio of greater
than 1 . 5, a short line-shaped dispense pattern will ensure that material reaches all corners. Void entrapment during place-
ment is minimized by ensuring a circular or oval-shaped dispense pattern with a steeply domed meniscus.
S ome recent work has shown that an alternate dispensing option for no-flow materials is to dispense a line of material adj a-
cent to the device to be underfilled. This material will then flow under the device via capillary action during the reflow pro-
cess. This is a proposed method of reducing voiding in no-flow underfilled devices.
8.3.2 Dispense Volume The volume of material dispensed will vary depending on the application. Only enough to under-
fill the package and to provide a necessary fillet is recommended. When using a no-flow underfill, too much material may
lead to ‘‘ floating’’ of the package after placement and a consequent loss of soldering yield (e. g. , opens). The proper amount
of underfill for a no-flow is dependent on the underfill density, the component/board gap, and the bump density.
8.3.3 Package Placement In some cases, packages may be placed on the dispensed underfill using the same parameters
as those without underfill. Package skew or ‘‘ floating’’ of the package at placement is most likely the result of too much
underfill. Lower placement velocity, higher placement force, and/or longer dwell time at placement may reduce the tendency
for the package to ‘‘ bounce’’ off the board after placement.
15
8.3.4 Application Problems Excess placement velocity or a poorly domed underfill deposit on the board may encourage
the entrapment of air under the substrate during package placement. When unnoticed until after cure, these voids are attrib-
uted to material volatility or substrate moisture. Void entrapment is minimized by ensuring a circular or oval-shaped dis-
pense pattern with a well-formed meniscus. Placement velocity must be low enough to ensure non-turbulent flow of mate-
rial outward from the center of the package site.
S harp or deep features (e. g. , unfilled vias) on the circuit board may also contribute to air entrapment during material appli-
cation. B ubbles can originate at open or untented vias, solder mask edges, uncovered traces, and other topologically differ-
ent features.
8.4 Flow Rate The dispense flow rate measurement and underfill flow rate measurement can be assessed utilizing tests that
have been developed by suppliers, users and industry sponsored efforts. Flow properties as the underfill is dispensed from
the syringe/needle at the substrate to board gap area and as the underfill flows under the package are both important from
the user’s perspective.
8.4.1 Dispense Flow Rate Measurement In order to characterize the flow behavior and determine shelf life/pot life sta-
bility of the underfill, a simple test involving a standard syringe size, needle type and geometry, set of specific pressure and
temperature conditions and the flow rate at the dispense station can be used for evaluation. Dispensing can be conducted in
an aluminum dish placed on a scale so that the amount (weight) of the underfill can be determined as a function of time.
This information may be recorded as weight per time or time per weight. This flow rate (weight per time) may correlate
with the underfill rate; however underfill rate relies on capillary action, and is influenced by chip size and standoff height.
The specific details of the testing conditions need to be coordinated with the underfill supplier.
8.4.2 Underfill Flow Rate The underfill flow rate or underfill time is one of the key attributes of an underfill encapsulant.
S uppliers typically use smooth or frosted glass slides with a predetermined gap established by shim stock to evaluate under-
fill times. In some cases, a fixture is used to maintain a stable gap and provide a more efficient (more than one slide assem-
bly can be prepared at a time) means of running the test.
The fixture is placed on the hot plate and monitored for the selected underfill temperature (typically 60-1 00 °C
[1 40-21 2 °F] ). When the selected temperature is reached, underfill material is dispensed at the gap area. The flow front is
timed to determine the time required to reach distances that approximate package dimensions. Information regarding under-
fill time and voiding predisposition can be obtained.
Other methods, such as glass slides set up with a variable gap that runs from ~0. 1 3 mm [~0. 005 in] to zero (a wedge shape)
can be used to expeditiously determine the compatibility of an underfill with various gap sizes. Another approach that may
be utilized is the use of a quartz bumped substrate that is consistent with the package size to be used in the application.
Different combinations of substrates with slides such as FR-4 can be evaluated. The predisposition to voiding can also be
assessed before and after cure. The glass slide/Quartz package tests are not designed to reflect underfill times in specific
applications, since solder masks, package size, bump count, bump spacing, bump location and surface cleanliness will all
have a dramatic effect on underfill times. This test may be used to compare relative performance for alternative underfill
suppliers, alternative processing temperatures, and shelf life/pot life of underfill encapsulants.
8.4.3 Flow Out and Bleed Flow out is the distance that the composite underfill flows away from the package. B leed refers
to underfill phase separation away from the package (see Figure 8-9) for capillary flow underfills. The polymer resin sys-
tem may flow outward (advancing contact angle) from the package leaving filler behind. B leed distance will exceed flow
out distance when phase separation occurs and some underfills will show this phenomenon because of the relatively smooth
and spherical filler.
The amount of flow out and bleed is measured by the distance that the resin advances. In some applications, it is important
to control the flow out/bleed to maximize the packaging density. There is not necessarily any reduction in reliability of the
underfilled component, however material that flows away from the package can cause problems by covering tooling holes,
fiducials, connectors, assembly pads and test points.
S ince there are a variety of underfills that have different viscosity and wettings capabilities, flow out/bleed can be optimized
through underfill materials selection. The dispensing rate, amount, pattern and substrate preheat temperature also can be
adj usted to help contain underfill in the assembly zone.
The most common way to measure the amount of the material is to use a scale to weigh the amount dispensed and adj ust
the line and pump speed to dispense a ‘‘ weight controlled line. ’’ This is used to setup the equipment and monitor it through
material changes.
16
Fi gu re 8-9 U n d erfi l l Bl eed
Piston pumps are commonly used to control the amount of material dispensed without direct impact of the material viscos-
ity changes over the useful working life of the material. Time/pressure dispensers often have too much shot-to-shot varia-
tion to provide consistent dispense volumes.
8.5 Spread/Slump If an underfill spreads or slumps too much it may flow too far from the device and encapsulate devices
not meant to be encapsulated with underfill.
A test method for evaluating spread/slump is:
a. Dispense at least 3 dots of adhesive 0. 65 cm [0. 256 in] diameter and 0. 25 mm [0. 01 0 in] thick on to a clean frosted
glass microscope slide (the dot pattern should be of uniform thickness. ).
b. Prepare two test specimens.
c. S pecimen No. 1 should be stored for 50 min to 70 min at 25 °C ± 5 °C [77 °F ± 9 °F] and 50 RH ± 25 RH.
d. S pecimen No. 2 should be cured as per supplier’s instructions.
e. Measure the percentage increase in adhesive dot diameter compared with the initial pattern.
When tested with this test method, the test pattern should not increase by more than 1 0% (or other agreed value) of the
printed dot diameter, and there should be no evidence of resin bleed. This level of spread insures minimal spread and a
decreased chance of the underfill interfering with adj acent devices.
8.6 Evaluation Methodology A number of destructive and nondestructive techniques are utilized to access the fill quality
of underfill encapsulant of a flip chip assembly. The quality of the fill is typically measured in terms of completeness of fill.
Appearance of voids, cracks in the assembly, and inhomogeneity of the underfill encapsulant (e. g. , channeling and resin/
filler phase separation) are defects. In a destructive mode, cross sectioning the assembly can yield information with regard
to encapsulant homogeneity and maj or voiding. In some cases, in particular with flexible circuits after cure, the component
can be peeled away and the fill removed to examine the package and board for void location, size, and frequency. An effec-
tive nondestructive method of imaging bumps, bump connections, and underfills, is known as acoustic micro imaging.
8.6.1 Acoustic Micro-Imaging CS AM is not necessarily feasible or practical for second-level interconnect, or stack
packages.
S canning Acoustic microscopes (S AM) utilize very high frequency ultrasound (typically 1 5 to 260 MHz) to non-
destructively produce images of internal features beneath the surface of a test obj ect. Frequencies in this range do not travel
through air; therefore water is used as a coupling medium. Defects such as delamination, cracking and voiding can be are
observed using S AM technology. Cracks in the encapsulant that fill with water cannot be detected with S AM since the
acoustic energy is not reflected from the air that is typically in a void or crack.
8.6.2 Destructive Tests for Voiding Unfortunately the presence of RF cans and multiple layers in the laminate and mul-
tiple interfaces in the CS P/B GA packages make it impossible to do the testing by acoustic microscopy for many underfilled
components.
17
Other test methods along with their advantages and disadvantages are shown in Table 8-1 :
Table 8-1 Test Methods

Method Advantages Disadvantages
Prying (w/ Cryanogenic Application) Fast, easy, can be done by anyone, checks Misses embedded voids that don’t reach
whole area under component at once the underfill/component interface
Finds buried voids, necessary Requires lab staff, only finds voids in the
Vertical Cross-Sectioning to inspect for filler settling particular plane sectioned, unless grind
further (very time consuming)
Horizontal / Planar Finds buried voids, checks Requires lab staff, very time consuming
Cross-Sectioning whole area under component
8.6.3 Assembly to Glass for Flow Visualization Although a package-to-glass assembly is not a perfect model, it can be
a powerful aid to understanding flow and problems with air entrapment or phase separation for capillary flow underfill. Any
bumped package can be bonded to the glass and commercial conductive adhesives are available for bonding. Epoxy-based
fluxes can also be used and flux can be applied from a fluxing drum or by pressing the bumps against a thin layer of flux
on a flat plate. A similar process using adhesive, commonly called ‘‘ polymer dip chip,’’ can be used for this assembly. Adhe-
sive is applied with a fluxing drum or coated onto a planar surface with a doctor blade or coating bar. The flux or adhesive
height should be less than the bump height. The bumped package is dipped into the flux or adhesive and removed with the
material clinging to the bumps. The package can now be placed onto the glass and hardened. The advantage of adhesive is
that it can be hardened at relatively low temperatures compared to the flux. This final assembly can be used to study under-
flow. Ideally, a camera is placed under the glass. However, since capillary flow (surface wetting) is a strong force, the flow
can be viewed through the glass (under a microscope) with the assembly upside down. This method allows assessment of
air entrapment caused by nonsymmetrical flow patterns, but will not always demonstrate phase separation dependent on sol-
der j oint widening or reduced gap.
8.7 Pot Life (In Dispenser) The dispense flow rate measurement method in 8. 4. 1 can be used to assess change. Underfill
encapsulant is typically checked for dispensing rate after 8 hours (one shift) at ambient conditions. A reduction of 50% in
the dispensing flow rate due to increased viscosity is generally considered the limit of pot life. The method in 8. 4. 2 (on
underfill flow rate) should be used to assess changes in underfill rate. Although viscosity can also be used to estimate pot
life, dispense and flow rate tests are much more meaningful. The user should evaluate the quality of underfilled assemblies
periodically to validate the amount of allowable reduction in pot life for a specific material.
Cure rate can also be tested using the method in 6. 2. 2 (gel time) although cure rate is rarely reduced for material kept at
ambient conditions in closed containers.
9 CURE PROCESS
9.1 Applied Life (After Dispensing) Underfill should be

cured soon after dispensing and flow out. The curing step
should occur as soon as possible after dispensing. However, in
A
situations where there is considerable time between dispens- B
ing and flow, the applied life test should be made. The under-
filled sample is cured after standing in the uncured state for
the maximum time anticipated between dispense and cure. A
IPC-030a-9-1
cross section should be examined for filler settling (see Figure
9-1 ). S ignificant settling, where there is a pronounced phase Figure 9-1 Filler Settling within Underfill
separation between resin and filler, is indicative of potential
A. Lower Filler Concentration
reliability problems. It is most common to start the curing
B. Higher Filler Concentration
process in the dispenser (on the post dispense heater station)
or in an inline oven j ust after dispense. Also, the amount of voiding in the cured underfill should be evaluated if there is
significant time between dispense and cure steps. Atmospheric moisture can be absorbed into the uncured underfill, and
cause voids during the curing process. Atmospheric moisture that is absorbed into the underfill can also affect the final prop-
erties of the cured underfill, such as Tg and CTE. Typically, absorbed moisture can lower the T g and also increase the dif-
ference between α 1
and α 2
CTE.
The gel time property of the materials (see 6. 2. 2) should be considered when addressing applied life of the dispensed mate-
rials. After the application process is complete, users may elect to gel the underfill to avoid long exposure to the atmosphere
18
and substrate that may inhibit curing. Certain epoxy chemistries may exhibit shorter gel times (indicating higher reactivity)
upon exposure to atmospheric moisture and will cure quickly. However, true cross linking will not occur, which results in
a lower T g value, poorer solvent resistance and reduced thermal stability.
9.2 Cure Process for Capillary Flow Underfill
9.2.1 Process Parameters The curing process will have a significant effect on final cured properties, especially the glass
transition temperature (Tg ) and the CTE. The process parameters that influence the cured properties include the rate of heat-
ing (ramp up) and cooling (ramp down), maximum temperature and time at temperature.
9.2.2 Cure Schedule Typically, cured properties listed on underfill data sheets are reported as total time at a given tem-
perature. Heat-up rates to this maximum temperature and cooling rates are not always specified in the cure schedules pro-
vided by the manufacturers. A common error that a user may make is to under cure an underfill because the user did not
consider warm-up time and thermal mass of the component/board. For example, an underfill that is specified to be cured for
3 0 minutes at 1 50 °C [3 02 °F] is used to underfill a device with a metal lid and soldered to a thick multi-layer board with
ground and power planes. If this PWB with underfill is placed into an oven at 1 50 °C [3 02 °F] for j ust 3 0 minutes it is
unlikely that the underfill saw 1 50 °C [3 02 °F] for a full 3 0 minutes. This would lead to an underfill that is under-cured and
can lead to increased probability of device failure.
For most materials, higher temperatures for longer periods produce higher Tg , lower CTE and greater adhesion (optimal
cure). In some cases, alternative cures schedules will be provided (e. g. ; lower temperatures, shorter duration, etc. ). Two-step
cures may also be used to reduce shrinkage and minimize stress for strain-sensitive applications. It is important to keep in
mind that as cure schedules are altered, final properties will vary. Curing at a lower temperature for a longer period of time
reduces stress but users should be sure they have achieved the necessary minimum temperature to initiate the curing pro-
cess. Also, alternative curing schedules can lead to changes in the final physical properties of the underfill.
9.2.3 Heating Rate This heating rate may influence the quality and reliability of the assembly. An acceptable heating rate
is influenced by the mechanical and physical properties of the package, board, and other components on the board as well
as the underfill material.
In general, slower heating rates may result in higher amounts of filler settling, but less warpage/stress on the assembly. A
fast heating rate (high ramp) for some underfill materials can cause one or more of the reactive monomers to volatilize and
each formulation must be tested for this phenomenon. Therefore, users need to be cognizant of the curing schedule utilized
and work closely with the underfill supplier to ensure the desired cured properties are attained.
9.2.4 Temperature Sensitivity When determining the cure profile (time and temperature) for an underfill system the effect
of that cure profile should be investigated on the rest of the assembly that is exposed to the heat excursion. For example,
components that are not rated to 1 60 °C [3 20 °F] can be damaged if the assembly that is going through an underfill cure
step at 1 65 °C [3 29 °F] . Therefore, the temperature rating of components and other portions of the assembly should be
assessed prior to implementing the underfill cure process.
9.3 Cure Process for No-Flow Underfill An IR or convection surface mount reflow oven can be used for curing the
no-flow/fluxing underfill. Reflowing the solder j oints before the underfill gels is a critical requirement for the reflow pro-
cess. If gelling of the underfill occurs before complete reflow of the solder, opens or malformed solder interconnects may
result.
Under-cure of the underfill during the reflow process can be addressed with a post-reflow bake. The assembly should be
kept reasonably dry between the two thermal events. Also, since the no-flow material is applied prior to the assembly and
components undergoing a heat cycle, trapped moisture can lead to increased voiding in the underfill. Therefore, a pre-drying
step is necessary to ensure substrates are dry and eliminate moisture related voiding.
One of the most significant differences between capillary flow underfills (applied after device reflow and cleaning) and no
flow underfills (applied before component placement and reflow) is the presence of organic acid constituents in the no flow
formulation to enable solder wetting to occur. Capillary underfills do not provide this function and so performance testing,
relative to this function, is not required. For the no flow underfills, however, it is necessary to establish acceptable, consis-
tent soldering performance.
The fluxing activity of the no flow polymer flux is a critical characteristic that largely determines its effectiveness in pro-
duction. Although all no flow polymer fluxes are designed to enable wetting, formulation characteristics and resulting shelf
19
and pot lives vary. As with any material used in production, users of these products should request specific guidance on stor-
age requirements and pot life expectations from the manufacturer. In addition, it is also important to request and/or deter-
mine material quality testing methods to assess the effectiveness of the material. Again, refer to the product manufacturer
for detailed guidance and information regarding specific formulations. A couple of simple, generic performance tests are
offered here for consideration.
The first simple test is a viscosity measurement. If the room temperature viscosity of the material differs from target values,
that viscosity shift might indicate that chemical reactions are proceeding at a rate significant enough to reduce the solder-
ing performance of the material.
Another simple test, involves testing the actual soldering performance of the flux. The recommended procedure is to place
a solder ball into a measured amount of the no flow underfill that has been applied to a wettable substrate. This test sample
is then subj ected to the required reflow profile. When the flux is performing optimally, the ball is expected to wet to the
substrate and spread to in a repeatable manner. No flow polymer flux suitability, relative to soldering performance, can be
j udged against the expected wetting performance.
9.4 Void Formation/Outgassing Outgassing in rigid underfill systems is a concern if there are solvents and/or diluents
present in the encapsulant. Moisture present in the substrate and board may also be a concern with regard to moisture out-
gassing during the curing process. Outgassing may also deposit material at critical surfaces such as contacts and connectors.
The visual detection of voids in the encapsulant after underfill, but prior to curing, may be minimal. However, after the cur-
ing process is complete, voids that may have been undetectable in the pre-cure condition may increase in size and become
more evident. Therefore, curing the specimen is important to assess void formation. While suppliers strive to provide under-
fill systems that will not form voids, the specific application and process conditions may exacerbate void formation. There-
fore, underfills must be tested in the end use application to ensure that there is minimal or no void formation.
Using the technique described in 8. 6. 2, an assessment of void formation can be conducted. Typically, the underfilled glass
assembly is examined before and after cure under a microscope to evaluate voiding formation. Quartz bumped packages may
also be used to assess void formation risk in a more representative test environment. To assess voiding in flip-chip on board
an acoustic micro imaging technique may be used, as described in 8. 6. 1 .
9.5 Cure Verification Curing of the underfill should be verified by comparing the fillet with a known cured sample. Pin
probing is commonly used to qualitatively test the hardness of cured underfill at full Tg of the underfill. S ome materials
change their appearances and/or optical characteristics once cured. For example, some underfill products have color reaction
cure indicators and these effects can be used as cure indicators.
The most accurate determination of reaction rates, cure kinetics and degree of cure is by thermal analysis such as Differen-
tial S canning Calorimitry (DS C) and Thermo-Mechanical Analysis (TMA). However, cured underfill is typically hard and
will not dent with pin probing. Furthermore, solvents, such as Isopropyl alcohol (IPA), should not remove more than a trace
of colorant.
TMA frequently requires a second scan in order to get consistent readings as the first scan tends to post cure the sample
resulting in misleading readings. DS C should be used to qualify degree of cure and can also determine T g . TMA can be
used to determine CTE and Tg of the fully cured material.
1 0 CURED UNDERFILL CHARACTERISTICS

1 0.1 Appearance The following applies to materials intended to have a fillet.
1 0.1 .1 Fillet Formation Fillets distribute stress and can increase reliability. Fillet height, volume, and spread vary due to
solder mask variations and flux residue. The lowest fillet usually occurs at the corners of a package. Fillet acceptance crite-
ria are covered in 1 3 . 4. 1 . The fillet should not cover the top of the package as it can make inspection difficult, interfere with
subsequent processing such as heat sink attach, and may also lead to increased stress and warpage. The highest stress points
in the fillet are at the corners of the package. Fillet formation can be controlled by the diameter of the needle, the proxim-
ity to the package and the amount and rate of material dispensed. Material over the package can cause unique stresses and
those adj acent components that wet to the underfill can cause unpredictable stresses.
1 0.1 .2 Color (Dye/Pigment) Most underfills are colored for visualization. Nonconductive dye is preferred over carbon
black that may reduce flow rate and possibly influence high frequency signal quality. B oth pigment and soluble dyes are
used.
20
1 0.2 Adhesion Direct determination of the bond strength

and/or the adhesion strength of the underfill layer to the sol-
A
der mask is a very difficult task. For comparative estimations
for the shear/adhesion strength of different materials, a sug-
gested test method is AS TM D1 002, S tandard Test Method for
Apparent S hear S trength of S ingle- Lap- Joint Adhes ively
B
B onded Metal S pecimens by Tension Loading (Metal- to-
Metal) (see Figure 1 0-1 ). Other methods include die-shear,
C
tensile pull, and peel tests. Reference www. AS TM. org for D
specific tests. IPC-030a-1 0-1
Underfill encapsulants exhibit some similarities in terms of fit, Figure 1 0-1 Underfill Adhesion Test
form, and function with conventional die attach materials that A. Stud C. Flux
have been used for decades. With this in mind, the adhesive B. Underfill D. Substrate
strength of the underfill may be evaluated as die attach mate-
rials have been tested over the years by means of shear and tensile testing. Peel testing may also be used on flexible sub-
strates. Lap shears may also be used to attain relative strength values.
Note: As the expected field failure is fatigue fracture (adhesive or cohesive) of the underfill layer, the obtained shear strength
values might not be good predictors for the material behavior. An accelerated life test (e. g. , drop test, bend test, vibration
test, thermal cycling, etc. ) that predicts the underfill life under field conditions is still the best way of estimating the underfill
material performance.
1 0.3 Shrinkage and Induced Stress A package under stress will typically show a slight bending that can be measured by
an interferometer, profilometer, or other sensitive methods. A perfect die/package under zero stress will have an infinite
radius of curvature. As stress is increased, the radius will be reduced to a measurable value.
1 0.4 Young’s Modulus Young’s modulus (modulus of elasticity) or stiffness is the ratio of stress to strain. The higher the
modulus value, the higher the rigidity of the material. A high modulus or rigid material is considered desirable for CTE sen-
sitive applications. For CSPs and other devices with interposers where the device failures are typically due to shock/drop or
vibration stress, a lower modulus material may be desired.
1 0.5 Coefficient of Thermal Expansion (CTE) Historically, when underfilling flip-chip devices, it is generally believed
that the CTE of the underfill should approximately match that of the interconnect system (i. e. , solder bump, adhesive j oint,
etc. ). The intent is for the underfill to expand and contract in unison with the interconnect material (an a1 range of 25 ppm/
°C - 3 0 ppm/°C is typical for tin/lead solders). However, there is continuing debate on the optimal CTE for underfill encap-
sulants. Conductive adhesives generally have higher CTE values than solder (50 ppm/°C - 60 ppm/°C) and may tolerate
underfills with similar CTE values.
The CTE above the glass transition temperature, designated as a2 , generally is much higher (3 00 - 400%) than a1 . As such,
underfills should be selected to have a Tg value well above the use temperature.
1 0.6 Glass Transition Temperature (Tg ) The desired T g has been typically determined by the maximum end use and/or
qualification upper most temperature exposure.
As noted in 1 0. 5, Tg is much less significant if expansion above and below the glass transition region is similar
(a1 ~ a2 ).
1 0.7 Chemical Stability As epoxy based material is the material of choice for the underfill layer, a high degree of chemi-
cal stability is expected. The chemical stability and activity of the underfill should be similar or superior to the encapsulant
materials used in electronic industry. In some cases, when some rework ability is required, a material that can be softened
at higher temperatures is used.
1 0.7.1 Determining Resistance to Solvents Prepare cured test specimens as indicated in 8. 5 and immerse them in sol-
vents listed in Table 1 0-1 for 1 5 minutes. Hang samples to dry for 1 0 minutes in normal ambient conditions. There shall be
no evidence of degradation in surface characteristics, such as chalking, tackiness, blistering, swelling, or color change when
examined visually. The user should test the cured adhesive for resistance to other solvents or chemicals likely to be encoun-
tered in this specific manufacturing process, during rework or in the final environment.
21
Table 1 0-1 Conditions for Chemical Resistance Testing

Chemical Thermal conditions
Isopropyl alcohol Ambient
Glycol-based cleaner Ambient
De-ionized water Ambient
1 0% alkaline detergent (pH 1 3) 60 °C ± 5 °C [1 40 °F ± 9 °]
1 0.8 Moisture Absorption Moisture absorption into cured underfill encapsulants can be an issue as absorbed moisture in
microvoids/craze sites in critical locations may impact device performance. In addition, the Tg , modulus, adhesion, dielec-
tric characteristics, and other properties may be reduced by moisture, thereby reducing high temperature exposure tolerance.
Underfill suppliers typically use a 24 h immersion in boiling water, measuring weight gain as a result. Not all suppliers
report moisture absorption values, but typical values fall in the 0. 3 % - 1 . 0 % range.
1 0.9 Hydrolytic Stability Hydrolytic stability refers to a polymer or liquid’s ability to maintain its cured or blended struc-
ture after exposure to elevated temperature and humidity conditions. S ome polymeric materials form a thermosetting struc-
ture when two components (e. g. , a resin and a hardener) are mixed. If an exposure to humid conditions causes the ‘‘ cured’’
material to separate into the two initial components, the material is said to ‘‘ revert. ’’ Hydrolytic stability tests are done to
make sure that cured polymeric structures do not revert in hot and humid conditions.
Historically, hydrolytic stability testing began in the 1 960s when two-part potting compounds in connectors were found to
revert to liquid components when exposed to j ungle conditions, often with disastrous results. Polymer science has progressed
a great deal in the last 40 years, and very few thermoset materials revert unless that is a specific design feature.
There are no nationally accepted or recognized test methods for the testing of hydrolytic stability specifically on underfill
materials. The general test approaches found in IPC-TM-650, methods or Telcordia (B ellcore) GR-78-CORE should be suf-
ficient to test the hydrolytic stability of an underfill material.
S imilarly, there are no nationally accepted criteria for the evaluation of an underfill material following hydrolytic stability
testing. In general, the underfill material should not exhibit degradation, cracking, chalking, or other evidence of attack or
reversion. The material should be able to survive the test exposure with little or no change in characteristics.
The surface should not be tacky. Apply a cotton-tipped swab to the surface of the tested material. No cotton fibers should
adhere to the surface.
The surfaces around the underfill material or under the underfill material should not show evidence of attack, metal migra-
tion, or corrosion.
1 0.1 0 Non-Nutrient In the past, polymeric materials often had fillers and additives made from food products (e. g.
starches). Unfortunately, when such materials were exposed to hot and humid j ungle conditions, biological agents (mold,
mildew) often fed on such additives, destroying the form, fit, and function of the equipment being ‘‘ protected. ’’ S ince that
time, most high performance materials require the use of polymer materials that will not ‘‘ feed’’ biological agents. With the
advances in polymer chemistry, very few modern materials in high performance electronics have any nutrient components.
The IPC has a test method (IPC-TM-650, TM 2. 6. 1 ) for determining if a polymeric material supports biological growth. This
test method involves taking a test substrate and exposing it (in controlled enclosures) to a mixture of biological spores, and
incubating the spores over a period of 28 days. If the polymer is non-nutrient, the agents will not attack it. While this method
applies specifically to solder masks and conformal coatings, the methodology and evaluation criteria can be applied to
underfill materials.
1 0.1 1 Surface Insulation Resistance S urface insulation resistance (S IR) is the measuring of the insulating characteristics
of a polymeric film, and how those characteristics change as a result of exposure to elevated temperature and humidity con-
ditions. The reader is referred to IPC-9201 , S urface Insulation Resistance Handbook, for a better understanding of this
method of test.
S IR can be used to evaluate underfill materials in the same manner as solder masks and conformal coatings. IPC-TM-650,
TM 2. 6. 3 . 1 and TM 2. 6. 3 . 2, refer to S IR (also called moisture insulation resistance) on polymeric materials.
Another method of testing polymeric materials (e. g. , masks, coatings, adhesives) can be found in Telcordia Technologies
GR-78-CORE, 1 3 . 2.
For both of these approaches, it is recommended that the candidate underfill be dispensed in rows across the selected comb
patterns (many to choose from), in a thickness similar to the intended end use thickness, and cured per the candidate cur-
ing process. B oth of the test approaches involve exposing the cured test sample to elevated temperature and humidity, for
4-7 days, with an applied electrical bias. Periodic resistance measurements are made.
22
It is desirable that the underfill material: maintain high resistance levels in humid conditions; do not cause or propagate cor-
rosion, and do not initiate or support electrochemical migration (dendritic growth). There are no current IPC standards for
acceptable S IR performance of underfill materials. All testing should be done with unprocessed controls for comparison pur-
poses.
1 0.1 2 Electrochemical Migration Resistance Electrochemical migration is very similar to S IR testing, but the exposure
times are generally much longer (e. g. , 7 days vs. 25). The test conditions, applied voltage levels, and evaluation criteria vary
depending on what specification is driving the testing.
IPC-TM-650, TM 2. 6. 1 4. 1 , and B ellcore GR-78-CORE, 1 3 . 1 . 3 , are identical. B oth involve a 596 hour exposure to 65 °C
[1 49 °F] / 85% RH environment, with a 1 0 volt DC applied bias, with measurements at 96 and 596 hours. The polymeric
material is j udged to be acceptable if no corrosion is observed, if no metal migration (more than 25% of the span of the
conductors) is observed, and if the resistance levels do not degrade more than one decade (a factor of 1 0) between the 96
and 596 hour readings.
1 0.1 3 Volume Resistivity Volume resistivity is a measure of the electrical insulation characteristics of a polymer material
through the bulk of the material (vs. across the surface). Volume resistivity of a polymeric material is most often measured
using AS TM D257, S tandard Test Methods for DC Resistance or Conductance of Insulating Materials.
Typically, AS TM D257 is used at lab ambient conditions (25 °C [77 °F] /50% RH) and a test voltage of 500 volts DC in
14 15
order to measure volume resistivity. Typical values range from 1 0 Ohm-cm to 1 0 Ohm-cm in the cured state; higher val-
ues yield better insulating properties. It is affected by: temperature, humidity, degree of cure, contamination, and voids or
cracks.
As with many other characterization tests in this section, the test can be run before and after a candidate environmental
exposure to determine if the environment caused a change in the volume resistivity.
1 0.1 4 Permittivity (Dielectric Constant) Dielectric constant (DC) and dissipation factor (DF) are inherent characteristics
of a polymeric material. DC and DF are important considerations as frequency increases, as they can affect the propagation
speed and loss of a radio frequency signal.
DC and DF are best visualized by analogy. Consider the amount of energy expended walking across a level putting green
on a golf course, compared to walking through the rough or through the woods. A high-speed material, such as Polytetra-
fluoroythelene (PTFE), has a low dielectric constant and a low dissipation factor (the putting green). Mainstream materials,
such as FR-4, have a higher constant and dissipation factor (the rough/woods).
Consequently, if an underfill material is used on or around a microwave (or other RF) signal path, the dielectric character-
istics of the material must be taken into consideration.
The most common method for measuring DC and DF of a polymeric material can be found in AS TM D1 50 S tandard Test
Methods for AC Loss Characteristics and Permittivity (Dielectric Constant) of S olid Electrical Insulation. If the method is
run as written, the DC and DF of the material is reported at 1 kilohertz (KHz) and 1 megahertz (MHz). This may not be
an accurate measure if you are using the polymeric material in the high MHz or gigahertz ranges. Keep this in mind when
looking at material data sheets. In addition, these tests are usually run at lab ambient conditions and on virgin material. This
may not be adequate if you need to know how the DC and DF change with environmental conditions.
Essentially, DC is the ratio of the capacitance of the insulating material to that of air (air=1 ). Typical values range from 3 -
4 in the cured state for underfills; lower values yield better insulating properties. It is affected by: temperature, humidity,
degree of cure, contamination, and voids or cracks.
DC and DF are properties and can be measured before and after a candidate environmental exposure (e. g. , humidity) to
determine if the exposure caused a shift in these material characteristics.
23
1 1 WORKMANSHIP
1 1 .1 Substrate Preparation The underfill dispense area on the board needs to be as flat as possible by design. Coatings
or surface preparation materials should be evaluated so that they do not interfere with the underfill material. Vias under the
package should be filled flush to the surface to prevent voiding. The dispensing side(s) should be given enough clearance
for needle movement. Flexible circuits should be rigidly supported under the assembly area.
1 1 .2 Cleaning Before Underfill Particulate, adhesive squeeze out (adhesive type flexible circuits), and removable flux
residues should be removed or minimized prior to the underfill process. The effects of contaminant on the underfill may fall
into the following categories:
a. Reduce the effective gap size under the package
b. Cause non-uniform filler distribution or filler filtering.
c. Flux residues may reduce the bonding strength of underfill to the relevant surfaces unless they are compatible. Flux resi-
due may also reduce the Tg of the cured underfill when mixed into the underfill.
1 1 .3 Cleaning After Cure Although, the underfill doesn’ t require cleaning after cure there may be other requirements that
require cleaning. The cured underfill material compatibility with any post cure cleaning process should be tested before both
processes are implemented.
1 2 PROCESS RELIABILITY ASSESSMENT

The function of underfill is to increase the reliability of the device that is underfilled. In order to be sure the correct mate-
rial is chosen, the appropriate reliability tests should be performed based on the end use environment. The two most com-
mon tests are thermal cycling and drop tests. In the case of thermal cycling, the underfilled device is thermally cycled while
the electrical connection is monitored for an open condition. Devices are tested to failure and a Wiebull distribution gener-
ated. In cases where the underfill is used to increase reliability for physical shock and vibration, then a drop test is com-
monly employed. In this case the underfilled device is dropped onto its side (edge) and also back (flat) and the device tested
for opens. Other tests include a) high temperature storage, thermal shock, three-point bending and vibration. The exact test
conditions (temperature ranges, etc. ) are best determined by the individual user and should be based on the product use con-
ditions as well as the adhesive material properties over the range of testing conditions.
1 2.1 Ionic Content Ions, especially chloride (Cl-) and sodium (Na+), can cause corrosion on pads and even printed cir-
cuit conductors. Typical limits for flip chip underfills are on the order of 5 ppm for each ion whereas CS P and B GA under-
fills are on the order of 3 0 ppm or higher. Ion Chromotography with extraction at 1 21 °C [250 °F] for 20 hours is used on
cured and pulverized samples of underfill to determine ion content. Ion content sensitivity will depend on the product func-
tion or use and material and device compatibility should be discussed with the material supplier.
1 2.2 Chemical Resistance The underfill material might offer some degree of protection for the package and the solder
interconnection when used under harsh environmental conditions. The expected protection has only limited value, i. e. , a
relatively large amount of contaminants are left on the surface by the no-clean flux process. A test may be required in order
to establish if the material solution is compatible with the environmental conditions expected in the field.
1 2.3 Mechanical Integrity The underfill material also provides a mechanical connection between the package and the
printed circuit board. Mechanical reliability improvement can be a key reason for adhesive usage for board level applica-
tions to mitigate stress due to shock, drop, and vibration. Excellent adhesion to both surfaces is critical for good mechani-
cal reliability. Underfill performance is highly sensitive to the adhesion of underfill to the solder mask and underfill/solder
mask adhesion should be tested to ensure mechanical integrity. In cases where the underfill material and solder mask have
strong adhesion, other physical properties such as elastic modulus dominate.
1 2.5 Post Soldering Processes (Capillary Underfill) As a general rule, capillary underfill is applied after the soldering/
reflow process.
As no cleaning is usually performed after the solder reflow, it is expected that the surfaces to which the underfill should
adhere will show a certain degree of contamination. Most of the contaminants are associated with the flux residue from the
no-clean fluxes. The compatibility between the flux and the underfill material is critical in order to achieve good adhesion
of the underfill layer.
24
In some cases, a solder reflow process might take place after the underfill layer was applied and cured. Under such circum-
stances, a possible failure mode may occur due to an uncontrolled expansion or melting (if thermoplastic materials are used)
of the underfill layer. If such manufacturing conditions might occur, a material that can withstand high-temperature and has
a high glass transition point should be selected.
1 2.6 Temperature Cycling (based on end use environment) The primary function of the underfill layer is to maintain the
mechanical integrity of the assembly over the useful life of the product and under the specified environmental conditions.
The mismatch in thermal expansion of the assembly in the absence of underfill is expected to allow for small displacements
between the package and board, due to the compliance of the solder balls which minimizes shear stresses that might destroy
the assembly. If the mismatch in the CTE between the board and the chip is small (for example: a silicon chip on ceramic
substrate), the elongation of the solder balls during the thermal cycle is small and the expected fatigue life of the assembly
is relatively long.
The existence of an underfill material also reduces the displacement of the substrate against the board and reduces the defor-
mation of the solder balls during the thermal cycle. A very rigid underfill will completely eliminate the problem of solder
fatigue. However, if the movement of the is eliminated, very large shear stresses can develop in some cases exceeding the
adhesion strength of the underfill material to the substrate surface and producing failure by delamination or component sol-
der interface fatigue. In order to avoid this type of failure, a proper selection of the underfill material, as a function of the
CTE, Glass Transition Temperature (Tg ), elongation should be performed (see Figure 1 2-1 ).
CG-2
7
6
5
4
0.9 CG-1 no-CG C 3
0.8 2
0.5 F CG-2 1
0
0.3 -40 -20 -0 20 40 60 80 1 00 1 20
A 0.2 E
0.1
0.05
300
0.03
0.02 250
0.01 200
CG-1
D 1 50
B 1 00 CG-2
50
0
-40 -20 -0 20 40 60 80 1 00 1 20
E IPC-030a-1 2-1
Figure 1 2-1 Material Property and Thermal Cycling Response

As in the case of two different materials with different CTEs that are j oined rigidly together, when the temperature changes,
the assembly bows in order to accommodate the mismatch in expansion between the two different material layers. This type
of deformation might produce axial stress in the solder balls that can produce the fracture of the ball, especially when the
assembly is exposed at low temperature. The axial stresses developed in the solder interconnect are functions of the rem-
nant stresses created during the underfill curing stage and the underfill material CTE. A correct selection of the material
properties will maintain the axial stress in the ball around zero or continuous compression.
25
1 3 OTHER CONSIDERATIONS
1 3.1 Reworkability When the cost of rework can be j ustified, reworkable underfill becomes a material of choice to enable
manufacturing flexibility. To ensure underfill reliability, the material must have good adhesion to both the board and the
component. However, for a material to be reworkable it needs to be easily removed and thus have poor adhesion. In order
to address the opposing adhesion needs for manufacturability and reliability, material suppliers have developed materials that
have good adhesion below the working temperature of electronic devices, but reduced adhesion at elevated temperatures
(such as above solder liquidus). Mechanical methods are typically used to rework underfilled devices as chemicals or sol-
vents that are effective in attacking the epoxy base resin of the underfill will also damage the board and surrounding devices.
A typical rework process uses a modified B GA/CS P rework station that has a collet that can heat the underfilled component
to be reworked. The component is heated above the liquidus of the solder and the T g of the underfill and a twisting motion
is used to break the adhesion between the underfill and the board. At this point any material remaining in the area where
the package was previously located can be removed by one of two methods. The first is a mechanical abrasion method,
where a brush (typically of natural fibers) attached to a rotatory mill tool is used to abrade the surface. The second method
is using a hot air vacuum tool or hot scrapping tool to ‘‘ peel’’ the underfill residue from the board so a new part may be
soldered. The appropriate rework method will depend on the material characteristics as well as the board features and the
user should contact the material supplier for the most suitable method. The user can also look at the procedures in
IPC-771 1 /7721 for conformal coating materials and determine if these procedures can be leveraged for the removal of the
underfill material. Once the pads and lands are cleaned, a new device can be attached and underfilled. The maximum num-
ber of rework cycles must be agreed upon by the manufacturer and the customer along with rework quality criteria.
Common failure mechanisms that can be induced during the rework process are:
a. Removal of solder mask;
b. Removal of pads;
c. Removal of traces; and
d. Laminate damage
Not all of these failures will be visible after the removal of the underfill device (e. g. , damage to the interior of the laminate).
1 3.1 .1 Rework of Adjacent Components When reworking adj acent components, care should be taken to minimize the
temperature exposure on the underfilled device. If the solder encapsulated in the underfill reflows, it will expand and can
result in cracks in the underfill possibly leading to shorts, opens, or other damage.
1 3.2 Determination of Cure Although some underfills have been formulated such that their color may change during cure,
the best way to determine the degree of cure is to perform a DS C analysis on the specific underfill formulation. Any sig-
nificant exotherm detected during the DS C scan indicates that the underfill is not fully cured. The maj or cause of under-
cured underfill is the underfill itself has not been subj ected to the proper time and temperature. Users of the underfill have
to take into consideration the thermal mass of the part, board, and other items in the oven. Fifteen minutes in a stand-alone
oven at 1 65 °C [3 29 °F] is not necessarily equivalent to curing the underfill for fifteen minutes at 1 65 °C [3 29 °F] . In the
same manner that assemblers profile their solder reflow ovens, users of underfill should profile their underfill ovens to be
sure the underfill achieves the cure time and temperature necessary for complete cure. Test boards can be setup with ther-
mocouples embedded into the underfilled region to monitor the temperature of the underfill for different cure profiles to
optimize the cure process to ensure complete underfill cure.
1 3.3 Thermal Management Thermally conductive underfill may be required for heat dissipation purposes. Today’s com-
mon filler is spherical silica that does not provide significant thermal conductivity. More thermally conductive filler materi-
als, such as nitrides, can be made in both an approximately spherical shape and size required for underfill. Thermally con-
ductive underfills may become available if there is sufficient demand and the higher material cost is acceptable. However,
since the solder j oint is an efficient thermal conduit and array bump patterns can be designed for thermal piping underfills
are typically not formulated or utilized for thermal management. Heat sinks applied to the back (top) of the die/package rep-
resent a more widely leveraged heat management option.
26
1 3.4 Workmanship Standards for BGA and CSP Board Level Underfills The following section may be applicable to
visual criteria (of those things which can be visually observed up to 4X magnification for referee purposes), such as fillet-
ing or bond line examinations. Some portions of the criteria will require destructive analysis. Regular in-house sampling and
inspection plans should be followed.
All inspection criteria presented in this section should be verified through reliability testing of the PWB with the underfilled
package and/or the end device.
For package on package (PoP) devices whenever there is a reference to a ‘‘package’’ or ‘‘component body,’’ the entire PoP
device should be considered as one unit unless otherwise noted.
1 3.4.1 Visual Criteria
Target – Class 1 ,2,3

• Component underfill is complete with a good fillet all the way around; 50% fillet height up the side of the component
body; completely cured; no encroachment onto other components and no voiding whatsoever.
Acceptable – Class 1 ,2,3

• Encroachment onto leaded devices (except for unsealed connectors or components).
• Complete underfilling of neighboring array device (partial underfill would not be acceptable).
Acceptable – Class 1 ,2
• Underfill fillet should cover 90% of the perimeter of the package, and does not expose solder interconnects (see Figure
1 3-1 ).
• Underfill may encroach on the top of a package as long as it does not interfere with proper placement of a heatsink or
next higher level assembly.
Figure 1 3-1 Acceptable – Class 1 ,2,3

Acceptable – Class 3
• Fillet height extends up to 1 00% of the vertical side of the component body.
• Small pin holes and voids are permissible provided they do not exceed 5% of fillet surface area (95% perimeter) and do
not expose solder interconnects.
• Evidence of a positive wetting angle between the cured underfill and the bottom edge of the body of the component.
Copyright IPC-Association Connecting Electronics Industries 27

Defect – Class 1 ,2,3

• Underfill is completely missing from components where it
is required.
• Underfill has not flowed to surround all the peripheral sol-

der j oints (see Figure 1 3 -2 ‘‘ A’’ ).
• Uncured underfill material.
• Partial underfilling of a neighboring array device.

B
• Underfill that interferes with proper placement of a heat-
sink or next higher level assembly.
• After curing there should be no evidence of cracks within

the underfill fillet or delamination between the underfill and
device surface or board (no evidence of either cohesive or
adhesive separation).
• There should be no evidence of damage to the board or

device caused by the dispensing process (mechanical or A
thermal).
Defect – Class 1 ,2
Figure 1 3-2 Void Comparison
• A void only visible by planar cross-sectioning around the Yellow = acceptable voids.
perimeter of one or more OUTER ROW solder j oints. Red = unacceptable missing underfill
• Total void area is >1 0% of the component area (see Figure
A. Underfill has not flowed to surround all the peripheral
1 3 -2 ‘‘ B ’’ ).
solder joints
B. Total void area is >1 0% of the component area
Defect – Class 3
• No evidence of a positive wetting angle between the cured underfill and the bottom edge of the body of the component
• S mall pin holes and voids >5% of fillet surface area.
• Material extending on to the top of the B GA, CS P, or flip chip.
Note: Encroachment onto chip components may lead to certain reliability defects and should be evaluated on a case-by-case
basis.
1 3.4.2 Underfill Voiding Under Devices

In order to meet most reliability requirements voiding under flip chip devices is not recommended.
For WLCS P, CS P, and B GA devices the level of acceptable voiding is dependent on end use reliability,
Note: Two adj acent solder balls, not fully covered by underfilled, may bridge during subsequent reflow or thermal cycling
processes.
1 3.5 Destructive Inspection After a volume calculation is performed, the underfill flow and coverage is typically wit-
nessed by the flow-out on the non-dispensed edges of the package. As an option, an internal destructive analysis may be
performed as part of a process development for a new product or material on test hardware for examination of the under-
filled area. This analysis is performed using a flat area microsection technique whereby the device is horizontally micro-
sectioned through the device to expose the array of solder balls encapsulated by the underfill material. Other lapping tech-
niques may also be used to expose the solder ball array.
The exposed area array may be examined for voiding in the underfill in accordance with the criteria described in 1 3 . 4. 2.
28
Target – Class 1 ,2,3

• No voids present anywhere in the underfill layer (see Fig-
ure 1 3 -3 ).
• Underfill is uniformly cured and homogenous throughout.
1 3.6 General Overview of Voiding Conditions and Reliabil-

ity Impact/Factors Any voids that encircle two or more sol-
der j oints should be considered a defect if the device will be
subj ected to any additional solder reflow, such as the follow-
ing:
a. Voids between 2 or more solder bumps.
b. Voids that bridge between 2 solder bumps.
c. Voids that are larger than 1 0% of total area.
Other voiding conditions they may be acceptable dependent

on reliability testing and analysis, such as the following:
a. Voids that encircle one solder bump.
b. Voids that partially encapsulate a solder bump. Fi g u re 1 3-3 Targ et – Fl at Cross Secti on th at Sh ows N o
Voi d s i n th e U n d erfi l l Layer

c. Voids that do not contain any solder bumps.
Destructive testing can also be used to evaluate the uniformity of cure and homogeneity of the underfill. If evidence of
uncured material (see 6. 3 . 1 and 9. 5) is discovered:
• Reliability testing should be constructed to determine if this should be considered a defect for Classes 1 or 2.
• Considered a defect for Class 3 .
1 3.7 Workmanship Guidelines for Corner or Edge Bond

Underfills
Corner Glue Length – the longer of manufacturer‘ s guide-
lines which are typically 3 -6 solder balls deep down the pack-
age or 3 mm [0. 1 1 8 in] minimum on small packages.
Material should be dispensed only along the edges or corners

of the package (see Figure 1 3 -4), single dots can lead to a
high degree of stress and early fatigue failure.
Fillet Height Up Substrate – 50 - 1 00% up the package sub-

strate throughout the entire dispense length. Closer to 1 00% is
preferred.
Fillet Width – in the range of 1 to 1 . 5 times the total package

height or 1 mm [0. 03 9 in] minimum on packages of 1 mm
[0. 03 9 in] or less in package height (see Figure 1 3 -5).
Glue Shape – should be flat to convex throughout the length

Fi g u re 1 3-4 I m ag e of a Com pon en t wi th E d g ebon d
U n d erfi l l
of dispense. Concave is not acceptable.
A B
Fi g u re 1 3-5 Fi l l et Wi d th
A. 1 00% of packag e su bstrate h ei g h t C. Packag e h ei g h t
B. Fi l l et h ei g h t D. Fi l l et wi d th
29
1 4 TROUBLESHOOTING
1 4.1 Inadequate Flow
1 4.1 .1 Viscosity Measure the viscosity to make sure it is in the specified range and the syringe is within the material sup-
plier’s pot life recommendations. If the viscosity is higher than stated in the specifications, the material may be partially
polymerized, due to thermal exposure or age and may need to be disposed If the material quality is acceptable, verify that
the material is intended to flow at the desired dispense or flow rate. S ome materials have rates of less than 1 cm/min/80 °C
and dispense and flow conditions are specific to underfill formulations.
1 4.1 .2 Wetting Good flow requires good wetting and, in turn, requires that the surface tension of the underfill be lower
than that of the surface energies of the package or board to be wetted. If poor wetting occurs, check the surface for con-
taminants. For example, certain contaminants, e. g. , silicones, may lower the surface energies of the surfaces, causing non-
wetting or dewetting.
1 4.1 .3 Mechanical Blockage The most common reason for
A
inadequate flow is mechanical blockage by too small a gap,
dense bump pattern, excessive flux residue, too thick a solder
mask or some other means of restricting movement of the
filler. S ince the epoxy system can flow along extremely small
B
channels, it is the mineral filler that gets blocked. Phase sepa-
ration in s p ecific areas , s uch as after a b ump, s ug ges t
mechanical blockage (see Figure 1 4-1 ).
1 4.2 Phase Separation

C
First, make sure that phase separa-
tion of the solid filler from the underfill liquid is not due to
IPC-030a-1 4-1
blockage (see 8. 2. 2. 4). Very low viscosity systems (<5 Pa s)
may show phase separation if the path is rather long, e. g. ,
Figure 1 4-1 Filtering Phase Separation
A. Package
large die. This is a natural consequence of material science
B. Board
and a different underfill may be required. Underfill with a
C. Restriction
higher resin viscosity or smaller filler size will typically show
lower phase separation. Also note that a modest amount of
filler/resin separation does not necessarily mean reduced per-
formance.
1 4.3 Voids
1 4.3.1 Voids Before Cure If voids are detected in the ‘‘ wet’’ underfill the three most likely causes are air being trapped
by an un-optimized flow pattern, air bubbles in the material as packaged, or air being introduced from the dispensing
parameters. Dispense and examine underfill on a glass slide for bubbles to evaluate bubbles resulting from packaging or dis-
pense parameters. Examine flow out by the method described in 8. 6. 2 to evaluate void entrapment from un-optimized flow
patterns.
1 4.3.2 Voids After Cure Voids after cure can be due to trapped air. Check the uncured sample by the test method in 8. 6. 2
and look closely for microscopic voids. Heating during curing will increase void size due to thermal expansion. If no voids
are visible in uncured samples, the most common source is volatilized moisture from the package or board. The most com-
mon way to avoid this voiding defect is to pre-dry the assembly before underfilling. This pre-drying step is critical in the
application of no-flow or fluxing underfills since they typically have not undergone a previous reflow step to drive out the
moisture from the packages or boards. If the voiding is not moisture related, volatilization of underfill material is likely sus-
pect. A non-reactant (e. g. , solvent) may be present in the material that is producing the void due to an un-optimized cure
profile. A slower temperature ramp or a gel stage may be required to allow solvent to volatilize before the underfill mate-
rial gels or begins to cure. Finally, the voids could be due to unreacted flux outgassing during the underfill curing process.
In this case, the reflow process must be optimized to minimize the amount of volatile flux remaining on the assembly.
1 4.4 Inadequate Cure Most underfill will not cure to 1 00% since reactants become trapped in the cross-linked matrix as
the polymerization advances and assemblers are reluctant to cure above the Tg where there is sufficient molecular cure. The
30
adequately cured underfill should (1 ) be very hard, (2) not dented by probing, and (3 ) resist color rub-off by solvent. If the
material is not sufficiently cured, the most likely cause is either insufficient temperature and/or time. Material suppliers often
specify curing conditions without stipulating that time needs to be allotted for the assembly and underfill to ramp to the cure
temperature. The actual package and underfill temperature can be tested by inserting a small thermocouple into the underfill
region and running the assembly through the cure profile. If cure cannot be achieved with appropriate thermal exposure,
material compatibility should be evaluated with other assembly materials such as flux residues. Underfill catalyst inhibition
can occur in limited instances due to contamination by flux residue.
1 4.5 Poor Adhesion Most organic substrates, solder masks and underfills are made with high epoxy content and these
systems are usually compatible. Factors that reduce adhesion include:
a. The presence of contaminants, especially incompatible flux residue;
b. Moisture in the substrate; and
c. S pecific underfill is not suited for the assembly die/package interface.
Marginal adhesion can be improved by boosting the cure temperature. Improved adhesion, can be obtained by a prebake at
1 50 °C [3 02 °F] for 3 0 min. If interfacial adhesion remains a concern for the product, material compatibility between sol-
der mask and board level adhesive formulations should be discussed with the respective material suppliers.
1 4.6 Thermal Cycle Failure Thermal cycle failures with board level adhesives can occur due to multiple individual and/or
combined factors. S ome common causes of premature thermal cycle failure are:
a. Poor adhesion.
b. Moisture in assembly.
c. Voids.
d. Modulus of underfill is too high or too low.
e. Cure temperature peak and/or durations too low.
f. Excessive shrinkage of underfill (low Rc).
g. CTE Mismatch between the underfill and the package-solder-board stackup.
h. Thermal exposure of product exceeds thermal range of material (i. e. , outside material’s Tg)
Failure analysis should be done on the failing units along with supporting material characterization to support hypothesis
testing for the different failure modes present in the units.
31
AN SI /I PC-T-50 Term s an d Defi n i ti on s for
I n tercon n ecti n g an d Packag i n g E l ectron i c Ci rcu i ts
Defi n i ti on Su bm i ssi on /Approval Sh eet
The purpose of this form is to keep SUBMITTOR INFORMATION:

current with terms routinely used in Name:
the industry and their definitions.
Individuals or companies are Company:
invited to comment. Please
complete this form and return to: City:
IPC State/Zip:
3000 Lakeside Drive, Suite 309S Telephone:
Bannockburn, IL 6001 5-1 249
Fax: 847 61 5.71 05 Date:
❑ This is a N EW term and definition being submitted.
❑ This is an ADDI TI O N to an existing term and definition(s).
❑ This is a CH AN G E to an existing definition.
Term Defi n i ti on
If space not adequate, use reverse side or attach additional sheet(s).

Artwork: ❑ Not Applicable ❑ Required ❑ To be supplied
❑ Included: Electronic File Name:
Document(s) to which this term applies:
Committees affected by this term:
O ffi ce U se
I PC O ffi ce Com m i ttee 2 -30
Date Received: Date of Initial Review:

Comments Collated: Comment Resolution:
Returned for Action: Committee Action: ❑ Accepted ❑ Rejected
Revision Inclusion: ❑ Accept Modify
I E C Cl assi fi cati on
Classification Code • Serial Number

Terms and Definition Committee Final Approval Authorization:
Committee 2-30 has approved the above term for release in the next revision.
Name: Committee: IPC 2-30 Date:
Put IPC MEMBERSHIP
Choose the Best Membership Option for Your Company
With a growing portfolio of member benefts and products and services, IPC provides a number
of membership options that deliver signifcant savings on key products and services. Whether
you prefer IPC’s Classic Membership, want the added focus of our Technology or Business
Packages, or need the comprehensive Business & Technology Package, IPC has a membership
option to suit your company’s needs.
Classic Membership
IPC’s Classic Membership provides core benefts to all employees at the company site or facility:
• 24/7 online access to members-only resources, including original articles and
presentations on the latest technical issues and industry/market trends
• One single-user download of each new or revised IPC standard within 90 days of
publication (with approximately 50 standards documents developed annually, that
represents an average savings of more than $2400 each year)
• 50% discount on IPC standards
• Signifcant discounts on IPC publications and training materials
• Signifcant discount on exhibiting at IPC events, including IPC APEX EXPO
• Reduced registration rates on IPC conferences and other educational events
• Access to participate in IPC market research studies (along with a complimentary report
for each study in which company participates)
Business Package
O fering all of the benefts of IPC’s Classic Membership, the Business Package Membership
provides additional benefts valued at more than $2,500:
• One All Access Package registration to IPC APEX EXPO®
• One registration to IMPACT 2014: IPC on Capitol Hill event
• One registration to an IPC management/business conference
Why upgrade now?
• One copy of a Fast Facts market research study When you upgrade your
membership to the Business
Technology Package Package, Technology
O fering all of the benefts of IPC’s Classic Membership, the Technology Package Membership Package, or Business &
provides additional benefts valued at more than $2,500: Technology Package, you
• One All Access Package registration to IPC APEX EXPO® receive complimentary
• Your choice of one registration to IPC TechSummit or two registrations to other IPC access to key IPC events and
technical conferences relevant technology or market
• One copy of information that can help you
IPC International Technology Roadmap for Electronic Interconnections
• Two registrations to IPC technology webinars
advance your business goals
and enhance competitiveness
Business & Technology Package — all at a savings of more
Get the most value from your IPC membership with the Business & Technology Package that than $1,000 compared to à la
builds on IPC’s Classic Membership, plus: carte pricing. Plus, because
• Two All Access Package registrations to IPC APEX EXPO events are bundled into your
• Your choice of one registration to IPC TechSummit or two registrations to other IPC
membership, you can enjoy the
technical conferences
convenience of going through
the budgetary approval process
• One registration to IMPACT 2014: IPC on Capitol Hill event only one time instead of
• One copy of IPC International Technology Roadmap for Electronic Interconnections several times a year.
• One registration to an IPC management/business conference Even if your membership
• Two registrations to IPC technology webinars anniversary is months away,
• One copy of a Fast Facts market research study you can take advantage of
Put the resources of the entire industry behind your company by joining the added value of IPC’s new
IPC today! Learn more about IPC membership and apply online at membership packages today!
www.ipc.org/membership. Your current membership will
be prorated and applied to your
new membership package.
to WORK for your COMPANY
Use this helpful chart to compare the features of I PC ’s membership bundles; then select the best one for your company.
Features Classic Technology Business Business &

Membership Package Package Technology
Package
24/7 online access to members-only resources, x x x x
including original articles and presentations on
the latest technical issues and industry/market
trends
One single-user download of each new or revised x x x x
IPC standard within 90 days of publication
(with approximately 50 standards documents
developed annually, that represents an average
savings of more than $2,400 each year)
50% discount on IPC standards x x x x
Signi f cant discounts on IPC publications and x x x x
training materials
Signi f cant discount on exhibiting at IPC events, x x x x
including IPC APEX EXPO
Reduced registration rates on IPC conferences x x x x
and other educational events
Access to participate in IPC market research x x x x
studies (along with complimentary report for
each study in which company participates)
One All Access Package registration to IPC APEX x x Includes two
EXPO registrations
Your choice of one registration to IPC x x
TechSummit or two registrations to other IPC
technical conferences
One registration to IMPACT 201 4: IPC on Capitol x x
Hill event
One copy of IPC International Technology x x
Roadmap for Electronic Interconnections
One registration to an IPC management/business x x

conference
Two registrations to IPC technology webinars x x
One copy of a Fast Facts market research study x x
For more information about IPC’s membership options and packages, visit www.ipc.org/membership or contact the
Member Success team at membership@ipc.org .

Association Connecting Electronics Industries
Application for Site Membership

®
Thank you for your decision to join IPC. Membership is site specific , which means that IPC member benefits are available to
all individuals employed at the site designated on this application.
Compa ny N a me
Street Ad d ress
City Sta te Zip/Posta l Cod e Cou ntry
M a in Ph on e N o. M a in Fa x
Compa n y E-ma il a d d ress Website U RL
N umber of Employees a t this Site N umber of Employees Corpora te-Wid e
N a me of Prima ry Con ta ct
Titl e M a il Stop
Ph on e Fa x E-ma il
To best serve your specific needs, please indicate the most appropriate member category for your facility. (Check one box only.)
❏ Printed Circuit Board Manufacturer

Facility manufactures and sells printed circuit boards (PCBs) or other electronic interconnection products to other companies.
What products do you make for sale? (check all that apply)
o One and two-sided rigid, multilayer printed boards o Flexible printed boards o Other interconnections
o Printed electronics
___________________________________________________________________________________________________________
❏ Electronics Manufacturing Services (EMS) Company
Facility manufactures printed circuit assemblies, on a contract basis, and may offer other electronic interconnection products for sale.
__________________________________________________________________________________________________
❏ OEM — Original Equipment Manufacturer
Facility purchases, uses and/or manufactures printed circuit boards or other interconnection products for use in a final product,
which we manufacture and sell.
What is your company's primary product line? ____________________________________________________________________
❏ Industry Supplier
Facility supplies raw materials, equipment or services used in the manufacture or assembly of electronic products.
Which industry segment(s) do you supply? ❏ PCB ❏ EM S ❏ Both ❏ Prin ted elec tronics
What products do you supply? _________________________________________________________________________________
❏ Government, Academia, Nonpro f t

Organization is a government agency, university, college or technical or nonprofit institution which is directly concerned with design,
research and utilization of electronic interconnection devices.
❏ Consulting Firm
What services does the firm provide? ___________________________________________________________________________
Application for Site Membership
®
Membership Packages and Dues

Membership will begin the day the application and dues payment are received, and will continue for one or two years (savings of 1 0%) based on the choice
indicated below. All fees are quoted in U.S. dollars.
Please check one:
Classic Membership Business Package Technology Package Business & Technology
Membership Membership Package Membership
One year Two years One year Two years One year Two years One year Two years
Primary Facility/site: The f rst site of an organization to
join IPC membership
o $1 ,1 00 o $1 ,980 o $2,600 o $4,680 o $2,600 o $4,680 o $4,000 o $7,200
Additional Facility/site: Membership for a facility of
an organization that has a different location than its o $900 o $1 ,620 o $2,400 o $4,320 o $2,400 o $4,320 o $3,800 o $6,800
Primary Facility membership
Company with annual revenue of less than $5 million o $650 o $1 ,1 70
Government agency, academic institution or nonpro f t
organization
o $300 o $540
Consulting f rm (employing fewer than six individuals) o $650 o $1 ,1 70
Explanation of Packages
Classic Membership — IPC’s classic membership provides core benefits to all employees at a company site/facility:
• 24/7 online access to members-only resources, including original articles and presentations on the latest technical issues and industry/market trends
• One single-user download of each new or revised IPC standard within 90 days of publication (with approximately 50 standards documents developed
annually, that represents an average savings of more than $2400 each year)
• 50% discount on IPC standards
• Significant discounts on IPC publications and training materials
• Significant discount on exhibiting at IPC events, including the industry’s flagship event IPC APEX EXPO ®
• Reduced registration rates on IPC conferences and other educational events
• Access to participate in IPC market research studies (along with a complimentary report for each study in which company participates)
Business Package — Offering all of the benefits of IPC’s Classic Membership, the Business Package Membership provides additional benefits
Membership valued at more than $2,500:
• One All Access Package registration to IPC APEX EXPO • One registration to an IPC management/business conference
• One registration to IMPACT 201 4: IPC on Capitol Hill event • One copy of a Fast Facts market research study
Technology Package — Offering all of the benefits of IPC’s Classic Membership, the Technology Package Membership provides additional benefits
Membership valued at more than $2,500:
• One All Access Package registration to IPC APEX EXPO • One copy of IPC International Technology Roadmap for Electronic
• Your choice of one registration to I PC TechSummit or two registrations to Interconnections
other IPC technical conferences • Two registrations to IPC technology webinars
Business & Technology Package Membership — Get the most value from your IPC membership with the Business & Technology Package that builds on IPC’s
Classic Membership, plus all the benefits of the Business and Technology Memberships combined.
Payment Information (Purchase orders not accepted as a form of payment)

Enclosed is a check for $__________ Bill credit card: (check one)
❏ MasterCard ❏ American Express ❏ Visa ❏ Diners Club
Ca rd N o. Expira tion D a te Secu rity Cod e
Authorized Sig n a ture Please attach business card

Mail application with check *Fax/Mail application with credit card payment to: of primary contact here
or money order to: IPC
IPC 3000 Lakeside Drive, Suite 309 S
3491 Eagle Way Bannockburn, IL 6001 5
Chicago, I L 60678-1 349 Tel: +1 847-61 5-71 00
www.ipc.org Fax: +1 847-61 5-71 05

*Overnight deliveries to this address only.
Contact membership@ipc.org for wire transfer details. 01 /1 4
®
Standard Improvement Form IPC J-STD-030A

The purpose of this form is to provide the Individuals or companies are invited to If you can provide input, please complete
Technical Committee of IPC with input submit comments to IPC. All comments this form and return to:
from the industry regarding usage of will be collected and dispersed to the IPC
the subj ect standard. appropriate committee(s). 3 000 Lakeside Drive, S uite 3 09S
B annockburn, IL 6001 5-1 249
Fax: 847 61 5. 71 05
E-mail: answers@ ipc. org
www. ipc. org/standards-comment
1 . I recommend changes to the following:
Requirement, paragraph number
Test Method number , paragraph number
The referenced paragraph number has proven to be:
Unclear Too Rigid In Error
Other
2. Recommendations for correction:
3 . Other suggestions for document improvement:
S ubmitted by:
Name Telephone
Company E-mail
Address
City/S tate/Zip Date

3 0 0 0 La ke s i d e D ri ve , S u i te 3 0 9 S
B a n n o c kb u rn , I L 6 0 0 1 5
8 47 - 6 1 5 - 7 1 0 0 te l
8 47 - 6 1 5 - 7 1 0 5 fa x
www. i p c . o rg

I SBN #97 8- 1 - 61 1 93 - 1 3 5 - 8

Ipc J-Std-030a-2014

Uploaded by

Copyright:

Available Formats

Ipc J-Std-030a-2014

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ipc J-Std-030a-2014

Uploaded by

Copyright:

Available Formats

I PC J -STD-030A

Copyright IPC-Association Connecting Electronics Industries

Standards Should: Standards Should Not:

Thank you for your continued support.

Selection and Application

Developed by the Underﬁll Materials Design, Selection and Process Task

Supersedes: Users of this publication are encouraged to participate in the

3 000 Lakeside Drive, S uite 3 09S

B annockburn, IL 6001 5-1 249

Phone ( 847) 61 5-71 00

Fax (847) 61 5-71 05

Copyright IPC-Association Connecting Electronics Industries

Copyright IPC-Association Connecting Electronics Industries

Assembly and Joining Underfill Materials Design, Technical Liaison of the

Vice Chair B rian J. Toleno

Underfill Materials Design, Selection and Process Task Group

Constantino Gonzalez, ACME Corporation David Hillman, Rockwell Collins

This Page Intentionally Left Blank

1 .2 Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7. 4 Pot Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1

1 .3 Deﬁnition of Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7. 4. 1 Viscosity Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1

7. 4. 2 Flow Rate Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1

3 TE RM S AN D DE FI N I TI O N S ...................................... 2 8. 2. 1 Dispensing Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3

4. 2. 5 Corner Glue/Corner Tack Underﬁlls 8. 3 . 1 Dispensing Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5

(Applied After S MT Reﬂow) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 8. 3 . 2 Dispense Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5

5. 1 Footprint Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 8. 3 . 4 Application Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 6

5. 2 Gap S ize . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 8. 4 Flow Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 6

5. 3 Pad Redistribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 8. 4. 1 Dispense Flow Rate Measurement . . . . . . . . . . . . . . . . . . 1 6

8. 4. 2 Underﬁll Flow Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 6

6. 2. 2 Gel Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 9 CU R E PRO CE SS .................................................... 1 8

6. 3 Material Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 9. 1 Applied Life (After Dispensing) . . . . . . . . . . . . . . . . . . . . . . 1 8

6. 3 . 1 Flux Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 9. 2 Cure Process for Capillary Flow Underﬁll . . . . . . 1 9

6. 3 . 2 B oard S urface Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 9. 2. 1 Process Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 9

6. 3 . 3 Component S urface Compatibility . . . . . . . . . . . . . . . . . . . . . 9 9. 2. 2 Cure S chedule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 9

6. 4 Alpha Particle Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 9. 2. 3 Heating Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 9

7. 1 Packaging and Containers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0 9. 4 Void Formation/Outgassing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

7. 1 . 1 Voids/B ubbles in Packed Material . . . . . . . . . . . . . . . . . . . 1 0 9. 5 Cure Veriﬁcation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

1 0 CU RE D U N DER FI LL CH ARACTER I STI CS . . . . . . . . . 20 1 3.6 General Overview of Voiding Conditions

1 0. 1 Appearance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 and Reliability Impact/Factors . . . . . . . . . . . . . . . . . . . . . . . . . 29

1 0. 1 . 1 Fillet Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 1 3.7 Workmanship Guidelines for Corner or

1 0. 2 Adhesion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 1 4 TRO U B LESH OOTI N G .......................................... 30

1 0. 3 S hrinkage and Induced S tress . . . . . . . . . . . . . . . . . . . . . . . . . . 21 1 4. 1 Inadequate Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0

1 0. 4 Young’s Modulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 1 4. 1 . 1 Viscosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0

1 0. 5 Coefficient of Thermal Expansion (CTE) . . . . . . . . 21 1 4. 1 . 2 Wetting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0

1 0. 6 Glass Transition Temperature (Tg ) . . . . . . . . . . . . . . . . . . . 21 1 4. 1 . 3 Mechanical B lockage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0

1 0. 7 Chemical S tability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 1 4. 2 Phase S eparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0

1 0. 7. 1 Determining Resistance to S olvents . . . . . . . . . . . . . . . . 21 1 4. 3 Voids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0

1 0. 8 Moisture Absorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 1 4. 3 . 1 Voids B efore Cure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0

1 0. 9 Hydrolytic S tability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 1 4. 3 . 2 Voids After Cure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0

1 0. 1 0 Non-Nutrient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 1 4. 4 Inadequate Cure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0

1 0. 1 1 S urface Insulation Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . 22 1 4. 5 Poor Adhesion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1

1 0. 1 2 Electrochemical Migration Resistance . . . . . . . . . . . . . 23 1 4. 6 Thermal Cycle Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1

1 0. 1 4 Permittivity (Dielectric Constant) . . . . . . . . . . . . . . . . . . . . 23

Figure 8-7 Dynamic Phase Separation ............................. 15

Table 4-1 Guidelines for Evaluation .................................... 4

This Page Intentionally Left Blank

Selection and Application of

increase the life of the assembly solder j oints.

Types of underﬁll materials currently available in the market include:

• Capillary Flow Underﬁll